Anda di halaman 1dari 108

_________________________________

GIIGNL
_________________________________


LNG CUSTODY TRANSFER


HANDBOOK

FOURTH EDITION version4.00










GIIGNL2015



GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

DOCUMENT STATUS AND PURPOSE

Thisfourth (2015)editionof the GIIGNL LNGCustody TransferHandbookreflectsGIIGNLs understanding of


bestcurrentpracticeatthetimeofpublication.

Thepurposeofthishandbookistoserveasareferencemanualtoassistreaderstounderstandtheprocedures
and equipment available to and used by the members of GIIGNL* to determine the energy quantity of LNG
transferredbetweenLNGshipsandLNGterminals.Itisneitherastandardnoraspecification.

ThishandbookisnotintendedtoprovidethereaderwithadetailedLNGshipshorecustodytransferprocedure
as such, but sets out the practical issues and requirements to guide and facilitate a skilled operator team to
workoutasuitableprocedureforaspecificLNGshipshorecustodytransferapplication.

Due to the rapidly changing marketconditions, new (commercial) opportunities arise leading to new technical
solutionsanddifferentoperations(suchaspartialunloading,reloadingatanLNGimportterminal,shiptoship
LNGtransferoperations,developmentofthesmallscaleLNGmarketnotonlywithmuchsmallerLNGshipsand
in much smaller quantities, but also with a different ship design and other cargo containment systems), the
GIIGNLGeneralAssemblyrequestedtheGIIGNLTechnicalStudyGrouptoreviewandupdatethethirdedition
(version3.01March2011)oftheGIIGNLLNGCustodyTransferHandbook.Therefore,thisfourtheditionmust
beseenasanextensiverevisionofthepreviousversionincludingnotonlycustodytransferfromalargescale
LNGshiptoanonshoreterminal(orviceversa),butalsocontainingalotmoreusefulinformationwithregardto
shiptoshipLNGtransferandcustodytransferfromanonshoreterminaltosmallLNGcarriers.Howeveritisnot
specificallyintendedtoworkoutproceduresforoverlandLNGcustodytransferoperationsinvolvingLNGtrucks,
containers or trains, or for small scale LNG transfer such as bunkering or refuelling of ships and trucks. For
these,kindreferenceismadetotheGIIGNLRetailLNG/LNGasafuelhandbook.

Noproprietaryprocedure,norparticularmanufactureofequipment,isrecommendedorimpliedsuitableforany
specificpurposeinthishandbook.Readersshouldensurethattheyareinpossessionofthelatestinformation,
standardsandspecificationsforanyproceduresandequipmenttheyintendtoemploy.

GIIGNL, and any of its members, disclaims any direct or indirect liability as to information contained in this
documentforanyindustrial,commercialorotherusewhatsoever.

Thislatestversionreplacesallpreviouseditionsofthecustodytransferhandbook.Pleasealwaysconsultthe
GIIGNLwebsitewww.giignl.orgtocheckforthelatestversionofthishandbook,esp.whenreferringtoapdf
downloadoraprintoutofthishandbook.

*
GIIGNL GroupeInternationaldesImportateursdeGazNaturelLiqufiParis

(InternationalGroupofLiquefiedNaturalGasImportersParis,France)


GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

CONTENTS

1. INTRODUCTION 5 3.2.2. Timingofthelevelmeasurement 16


2. GENERALDESCRIPTIONOFTHE 3.2.3. Readings 16
MEASUREMENT 6
3.2.3.1 Readingofthelevelwithfloat
2.1. GENERALFORMULAFORCALCULATING gauges 16
THELNGENERGYTRANSFERRED 7
3.2.3.2 Readingofthelevelwith
2.2. GENERALSCHEMEOFTHE capacitanceandradargauges 17
MEASUREMENTOPERATIONS 7
3.2.4. Correctionofreadings 17
2.2.1. LNGVolume 7
3.2.4.1 Floatgauge 17
2.2.2. LNGDensity 8
3.2.4.2 Capacitance,radarandlaser
2.2.3. LNGGrosscalorificvalue 8 gauges 17
2.2.4. Energyofthegasdisplacedbythe 3.2.5. Useofsparelevelgauge 21
transferofLNG 8
3.2.6. Completeunloading(tankstripping) 21
2.2.5. Flowchartfordeterminingtheenergy
transferred 9 3.2.7. Automatedsystems 21

2.3. INSTRUMENTSUSED 9 3.3. CALCULATIONOFTHEVOLUMEOFLNG


TRANSFERRED 22
2.3.1. ForthedeterminationoftheLNGvolume 9
4. TEMPERATUREMEASUREMENT 23
2.3.2. ForthedeterminationofLNGdensityand
grosscalorificvalue 9 4.1. LIQUIDTEMPERATURE 23

2.3.3. Forthedeterminationoftheenergyof 4.1.1. Device 23


displacedgas 10 4.1.2. Testingandaccuracy 24
2.3.4. Forthedeterminationoftheenergyof 4.2. VAPOURTEMPERATURE 24
Gastoengineroom 10
5. VAPOURPRESSUREMEASUREMENT 24
2.3.5. Periodicinstrumentsrecalibration 10
6. SAMPLINGOFLNG 26
2.4. STANDARDIZATION 10
6.1. LNGSAMPLINGPRINCIPLES 26
2.5. PARTIALLOADINGORUNLOADINGOF
LNGCARRIERS 10 6.2. SAMPLINGPOINT 27

2.6. Gassingupandcoolingdownoperations 11 6.3. SAMPLINGPROBES 28

2.6.1. Gassingupoperations 11 6.4. PIPINGARRANGEMENTBETWEEN


SAMPLINGPROBEANDVAPORIZER 29
2.6.2. Coolingdownoperations 11
6.5. LNGVAPORIZERANDCONTROL
3. VOLUMEMEASUREMENT 12 DEVICES 29
3.1. GAUGETABLES 12 6.5.1. Maindevices 29
3.1.1. Useofgaugetables 12 6.5.1.1 Electricalvaporizer 29
3.1.2. Correctiontables 13 6.5.1.2 Waterorsteamvaporizer 31
3.1.2.1 Correctionaccordingtothe 6.5.2. Auxiliaryvaporizationcontroldevices 31
conditionoftheLNGcarrier 13
6.5.2.1 OnregasifiedLNGoutlet 31
3.1.2.2 Correctionsaccordingtothe
temperaturesintheliquidand 6.5.2.2 OnLNGinlet 32
gaseousphases 13 6.5.2.3 Onheatingfluid(water,steamor
3.1.3. Approvalbyauthorities 13 electricity) 32

3.2. INSTRUMENTSANDMETHODSFOR 6.5.2.4 Safetydevices 32


MEASURINGTHELEVELOFLIQUIDIN 6.5.3. Operatingparameters 32
THELNGCARRIER'STANKS 15
6.5.4. Gassamplecollectionsystems 32
3.2.1. Mainliquidlevelgaugingdevices 15
6.6. GASSAMPLECONDITIONING 33
3.2.1.1 Electricalcapacitancetypelevel
gauge 15 6.6.1. Gassamplecontainers 33
3.2.1.2 Floattypelevelgauge 15 6.6.2. Directpipingtoagasanalyzer 33
3.2.1.3 Radar(microwave)typelevelgauge15 6.6.3. Examples 33
3.2.1.4 Lasertypelevelgauge 16 6.7. PERFORMANCEOFTHEDEVICES 36

1
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

6.8. SAMPLINGPROCEDURE 37 10.2.2.2 Examplesofchartsofbasic


physicalconstants 45
6.8.1. Samplingperiod 37
10.2.2.3 Exampleofcalculation 45
6.8.2. Samplingfrequency 37
11. ANALYSISREPORT 46
6.8.3. Purging 37
11.1. IDENTIFICATION 46
6.8.4. Samplingparameters 37
11.2. BASICDATA 46
6.8.5. Utilisationofgassamplecontainers 37
11.3. RESULTS 46
6.9. SPOTSAMPLINGDEVICE 37
12. ENERGYOFGASDISPLACEDOR
7. GASANALYSIS 37 CONSUMED 46
7.1. TYPEOFGASCHROMATOGRAPH 38 12.1. ENERGYOFGASDISPLACEDFROM
7.2. CALIBRATION 38 THELNGTANKS 46
7.2.1. Calibrationgas/workingstandard 38 12.2. ENERGYOFGASCONSUMEDASFUEL
BYTHELNGCARRIER 47
7.3. ENVIRONMENTFORAGAS
CHROMATOGRAPHICSYSTEM 38 13. ENERGYTRANSFERDETERMINATION 47
7.4. ANALYSISOFREGASIFIEDLNGAND 14. ENERGYTRANSFERMEASUREMENT 47
RETAINEDSAMPLES 39 15. UNCERTAINTYOFTHEENERGY
7.5. UNCERTAINTYOFGASANALYSIS 39 TRANSFERDETERMINATION 48
7.6. RAMANSPECTROSCOPY 39 15.1. VOLUME 48
7.7. IMPURITIES 40 15.1.1. CargoLiquidLines 48
7.7.1. General 40 15.2. DENSITY 49

7.7.2. Specificationsandmeasurementoftrace 15.3. GROSSCALORIFICVALUE 49


impuritiesinLNG 40 15.4. SAMPLINGANDVAPORIZATION 49
7.7.2.1 Carbondioxide 40 15.5. GASDISPLACED 49
7.7.2.2 Sulphur 40 15.6. GASCONSUMEDINENGINEROOM 49
7.7.2.2.1 Totalsulphur 41 15.7. COMBINEDSTANDARDUNCERTAINTY
7.7.2.2.2 Sulphurcomponents 41 ANDEXPANDEDUNCERTAINTYOFTHE
ENERGYTRANSFERDETERMINATION 49
7.7.2.3 Mercury 41
15.8. ROUNDINGOFNUMBERSAND
8. DATAPROCESSING&TREATMENT 41 COMMERCIALIMPACT 50
8.1. GENERAL 41 16. SHIPTOSHIPLNGTRANSFER
8.2. DATAHANDLING:QUALITY 41 OPERATIONS 50

8.2.1. Dataprocessing 42 17. SMALLSHIPTOSHORELNGOPERATIONS 51

8.2.2. Datatreatment 42 18. RELOADINGOPERATIONSIN


REGASIFICATIONTERMINALS 51
9. DENSITY 42
19. LNGSALESCONTRACTCUSTODY
9.1. GENERAL 42 TRANSFERCHECKLIST 52

9.2. DENSITYCALCULATIONMETHODS 42 ENCLOSURE1:CONVERSIONFACTORTABLE
9.3. REVISEDKLOSEKMcKINLEYMETHOD 43 FORENERGYUNITS 55

9.3.1. Limitsofthemethod 43 ENCLOSURE2:LNGANDNATURALGAS
CUSTODYTRANSFERMETHODS 56
9.3.2. Formula 43
ENCLOSURE3:OTHERRELEVANTSTANDARDS
9.3.3. ExampleofLNGdensitycalculation 43
ANDREFERENCES 59
10. GROSSCALORIFICVALUE 43
ENCLOSURE4:NOTESONTERMINOLOGY 60
10.1. GENERAL 43
LISTOFFIGURES 61
10.2. METHODOFDETERMINATIONOFTHE
GROSSCALORIFICVALUE 43 LISTOFTABLES 62

10.2.1. Determinationwiththehelpof REFERENCES 63
calorimeters 44
APPENDIX1:USEOFINLINEMEASUREMENT
10.2.2. DeterminationofGCVbycalculation 44
OFLNGFLOW 64
10.2.2.1 Examplesofformula 44

2
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

APPENDIX2:LASERTYPECARGOTANKLEVEL
GAUGE 65

APPENDIX3:SOMERECOMMENDATIONSFOR
PARTIALRELOADINGOFCARGOTANKSOFLNG
CARRIERS,WITHREGARDTOBOTHSAFETY
ANDCUSTODYTRANSFERISSUE 66

APPENDIX4:EXAMPLEOFGAUGETABLES 68

APPENDIX5:EXAMPLEOFVOLUME
CALCULATION 72

APPENDIX6:EXAMPLESOFDISCONTINUOUS
ANDCONTINUOUSSAMPLINGAND
VAPORIZATIONSYSTEMS 77

APPENDIX7:DIRECTINLINEANALYSISWITH
RAMANSPECTROSCOPY 80

APPENDIX8:GRUBBsTEST 82

APPENDIX9:TABLESFORLNGDENSITY
CALCULATIONACCORDINGTONBS 86

APPENDIX10:EXAMPLEOFLNGDENSITY
CALCULATION 88

APPENDIX11:EXAMPLEOFGCVCALCULATION 91

APPENDIX12:UNCERTAINTYCALCULATIONS 93

APPENDIX13:EXAMPLEOFCOMMERCIAL
IMPACTOFROUNDINGOFNUMBERS 96

APPENDIX14:TWOCASESTUDIES
ILLUSTRATINGPOTENTIALRISKSOF
STRATIFICATIONANDROLLOVERINLNGSHIPs
CARGOTANKS 97

APPENDIX15:EXAMPLEOFCUSTODY
TRANSFERDATALOGSONBOARDALNG
VESSEL 101

3
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted


1. INTRODUCTION
Followingthepublicationin1985bytheN.B.S.ofitsstudy AdvanticaTechnologiesLtd.(UK)
"LNG Measurement A User's Manual for Custody BGInternational(UK)
Transfer" [8], the Executive Committee of the GIIGNL CMSEnergyTrunklineLNGCompany(USA)
(Groupe International des Importateurs de Gaz Naturel Distrigas(B)
Liqufi) considered it would be useful to write a Enagas(E)
handbook, as simple and as practical as possible, aimed GazdeFrance(F)
at organizations involved in the measurement of the NigeriaLNG(NI)
energy transferred in the form of LNG in the context of a NKKK(JP)
LNGpurchaseandsalesagreement,whetherthissalebe OsakaGas(JP)
F.O.B. [Port of loading], D.E.S. or C.I.F. [Port of ReteGasItalia(I)
unloading]. SIGTTO(UK)
TokyoGas(JP)
DuringitssessionofOctober1987,theGeneralAssembly TractebelLNGNorthAmerica(USA)
ofGIIGNLdecidedthatthispracticalhandbookshouldbe
drawnupbyaStudyGroupcomprisingcompanies ofthe
GIIGNLandcoordinatedbyDistrigasS.A(B). THIRD EDITION, MARCH 2010

The methods described in this handbook could serve to Since the second edition, several new international
improve existing procedures. They could also be used in standards and revisions of existing international
purchaseandsalesagreementsfortheGIIGNLmembers standards related to the subject of this handbook, have
andserveasareferenceinnewimportagreements. been published or are forthcoming. Also, technologies
and best current practices evolved in this past period.
This handbook is based on the measurement methods Therefore,theGIIGNLGeneralAssemblyrequestedthe
mostusedbyGIIGNLmembers. GIIGNL Technical Study Group to revise the second
edition (October 2001) of the GIIGNL LNG Custody
Detailedtestsoftheapparatususedcanbefoundin"LNG Transfer Handbook, with the upcoming new ISO
MeasurementStudy"ofN.B.S.[8]. standardISO10976Measurementofcargoesonboard
LNG carriers, which will replace and supersede the
We wish to thank the companies BG (UK) Distrigas currentISO13397(1997)standarduponitspublication.
Boston (USA) Enagas (E) Kansai Electric Power Co
(JP) Snam (I) Tokyo Electric Power Co (JP) Tokyo Moreover this third edition of the handbook has been
Gas Co Ltd (JP) Ruhrgas (D) CMS Energy Trunkline updatedandrevisedasappropriatetostreamlineitwith
LNG (USA) for their cooperation in producing this neworrevisedinternationalstandardssuchasISO,EN
handbook, and more particularly Gaz de France for andotherstandards.Theseinclude:
drawingupSections6and7ofthishandbookandOsaka
Gas Co Ltd for cocoordinating the studies of the EN437:TestGasestestPressuresAppliance
Japanesecompanies. CategoriesEdition2003
ISO8943Refrigeratedlighthydrocarbonfluids
Sampling of liquefied natural gas Continuous
SECOND EDITION, OCTOBER 2001 andintermittentmethodsEdition2007
ISO 69746 Natural gas Determination of
FollowingthepublicationoftheISO13398:1997standard composition with defined uncertainty by gas
"LNGProcedureforcustodytransferonboardship",the chromatography Part 6: Determination of
GIIGNL General Assembly requested the GIIGNL Study hydrogen, helium, oxygen, nitrogen, carbon
Group to revise the original edition (March 1991) of this dioxide and C1 to C8 hydrocarbons using three
GIIGNL LNG Custody Transfer Handbook, particularly capillarycolumnsEdition2002
takingintoaccountthisnewISOstandard. ISO16664:GasanalysisHandlingofcalibration
gases and gas mixtures Guidelines Edition
All13sectionsoftheoriginaleditionhavebeenreviewed 2004
and updated where appropriate. The following sections ISO 4259: Petroleum products Determination
havebeenthoroughlyrevised: and application of precision data in relation to
methodsoftestEdition2006
2. Generaldescriptionofthemeasurement ISO/TR 24094: Analysis of natural gas
3. Volumemeasurement Validation methods for gaseous reference
6. SamplingofLNG materialsEdition2006
7. Gasanalysis ISO10723:NaturalgasPerformanceevaluation
foronlineanalyticalsystemsEdition2002
Moreover,anewsectionwasadded: ISO 63261: Natural gas Determination of
14. LNGSalescontractcustodytransferchecklist. Sulphur compounds Part 1: General
Worked out examples for LNG density and GCV have introductionEdition2007
beenrearrangedinAppendices1and2. ISO 6327: Gas analysis Determination of the
water dew point of natural gas Cooled surface
We wish to thank all companies and organizations and condensationhygrometersEdition2007
their delegates who together contributed to this second ISO 19739: Natural gas Determination of
edition,viz.(inalphabeticalorder): sulphur compounds using gas chromatography
Edition2004

5
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

ISO 122131 2 3: Natural gas Calculation of appropriate. Especially Sections 6 and 7 have been
compressionfactorEdition2006 adapted to streamline it with the current best practices.
ISO15112:Naturalgasenergydetermination To make this handbook more readable, most of the
Edition2007 examples have been replaced to the appendices.
ISO181321:Refrigeratedlighthydrocarbon Moreover, all sections with regard to the uncertainty of
fluidsGeneralrequirementsforautomaticlevel the energy determination have been moved to Section
gaugesPart1:Gaugesonboardshipscarrying 15andthefollowingsectionshavebeenadded:
liquefiedgasesEdition2006 2.6Gassingupandcoolingdownoperations
ISO181322:Refrigeratedlighthydrocarbon 15.8Roundingofnumbers
fluidsGeneralrequirementsforautomaticlevel 16.ShiptoshipLNGtransferoperations
gaugesPart2:Gaugesinrefrigeratedtype 17.SmallLNGshiptoshoretransferoperations
shoretanksEdition2008 18. Reloading operations in regasification
ISO/DIS28460:Petroleumandnaturalgas terminals
industriesInstallationandequipmentfor With regard to these new Sections 2.6, 16, 17 and 18,
liquefiednaturalgasShiptoshoreinterface theaimofthisGIIGNLLNGCustodyTransferHandbook
andportoperationsEdition2009 is to integrate the specific conditions for this special
operations of gassingup and cooling down, for shipto
All14sections,enclosuresandappendicesofthesecond ship LNG transfer, for small LNG shiptoshore transfer
edition have been exhaustively reviewed and updated (or vice versa) and for reloading operations at
whereappropriate. regasification terminals, but not to integrate the small
scale LNG transfer operations (such as bunkering or
Wewishtothankall20companiesandorganizationsand fuelling ofshipsandtrucks, and filling of LNGtrucksor
their delegates who together contributed to this third containers).
edition,inalphabeticalorder:
We wish to thank all companies and organisations and
ActOnLNGConsultingUK their delegates who together contributed to this fourth
BotasMarmaraEreglisiTurkey edition,inalphabeticalorder:
BPSunburyonThamesUK
DistrigasofMass.GDFSuezEverett(Boston)USA BotasMarmaraEreglisiTurkey
DragonLNGMilfordHavenUK BGGroupHoustonTexas
ElengyGDFSuezParisFrance EnagasSpain
EnagasSpain FluxysLNGZeebruggeBelgium
ExxonMobilHoustonTXUSA GasNaturalFenosaSpain
FluxysLNGZeebruggeBelgium GateTerminalRotterdamTheNetherlands
GasNaturalMadridSpain GDFSuezParisFrance
GLNobleDentonLoughboroughUK NationalGridGrainLNGUK
KogasSeoulSouthKorea RasGasRasLaffanQatar
NationalGridGrainLNGUK ShellGlobalSolutionsTheHagueTheNetherlands
OsakaGasOsakaJapan TokyoGasParisFrance
ShellGlobalSolutionsTheHagueTheNetherlands TotalParisFrance
RasGasRasLaffanQatar
RENAtlanticoSinesPortugal
SempraLNGSanDiegoCAUSA
SGSBelgium
2. GENERAL DESCRIPTION OF
SIGTTOLondonUK THE MEASUREMENT
TokyoGasTokyoJapan
TotalParisFrance
Accuracy

FOURTH EDITION, FEBRUARY 2015 Thetermmeasurementaccuracyisdefinedinthemost

recent version of the International Vocabulary of


Due to the rapidly changing market conditions, new
Metrology (JCGM_200:2012) as closeness of
(commercial) opportunities arise leading to new technical
agreementbetween a measured quantity valueand a
solutions and different operations (such as partial
true quantity value of a measurand. Measurement
unloading, reloading at an LNG import terminal, shipto
error is defined as measured quantity value minus a
ship LNG transfer operations, development of the small
reference quantity value.
scaleLNGmarket notonly with muchsmaller LNGships

and in much smaller quantities, but also with a different
Uncertainty, combined standard uncertainty and
ship design and other cargo containment systems), the
expanded uncertainty
GIIGNL General Assembly requested the GIIGNL

Technical Study Group to review and update the third
The terms measurement uncertainty, combined
edition (version 3.01 March 2011) of the GIIGNL LNG
standard uncertainty and expanded uncertainty (see
CustodyTransferHandbook.
Section15)areusedasdefinedintheJCGM100:2008

document:EvaluationofmeasurementdataGuideto
Furthermore there is a continuous evolution in the LNG
theexpressionofuncertaintyinmeasurement.
(sampling) technology and this fourth edition tries to

incorporateandbeinlinewithneworrevisedinternational

standards.

All sections, enclosures and appendices have been


thoroughly reviewed, amended and updated where

6
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

2.1. GENERAL FORMULA FOR


CALCULATING THE LNG ENERGY DLNG = thedensityof LNGloaded orunloaded
3
inkg/m .
TRANSFERRED
The formula for calculating the LNG transferred depends GCV LNG = the gross calorific value of the LNG
on the contractual sales conditions. These can relate to loaded orunloaded in MMBTU/kg.The
several types of sale contract as defined by Incoterms grosscalorificvalueisgenerallyusedin
2010. In case of rules for sea and inland waterway international LNG trading rather than
transport, the most commonly used are a FOB sale, a thenetcalorificvalue,seeEnclosure4.
CFRsaleoraCIFsale.
Egas displaced = the net energy of the displaced gas,
In the case of a FOB (Free On Board) sale, the alsoinMMBTU,whichiseither:
determination of the energy transferred and invoiced for sent back by the LNG carrier to
will be made in the loading port. There is another sale shore or to another LNG carrier
contract similar to FOB, named FAS (Free Alongside when loading (volume of gas in
Ship). cargo tanks displaced by same
volumeofloadedLNG),
InthecaseofaCIF(CostInsurance&Freight)oraCFR or,gasreceived bytheLNGcarrier
(Cost and Freight) sale, the energy transferred and initscargotankswhenunloadingin
invoicedforwillbedeterminedintheunloadingport. replacement of the volume of
dischargedLNG.
Other rules exist for any mode of transport which can
hence also apply for maritime transport such as DAT Egas to ER = if applicable, the energy of the gas
(DeliveredAtTerminal)andDAP(DeliveredAtPlace). consumed in the LNG carriers engine
room (also including all gas burnt by
InFOBcontracts,thebuyerisresponsibletoprovideand theshipsGCU(GasCombustionUnit))
maintain the custody transfer measurement systems on during the time between opening and
board the vessel for volume, temperature and pressure closing custody transfer surveys, i.e.
determinationandthesellerisresponsibletoprovideand used by the vessel during the LNG
maintain the custody transfer measurement systems at transferoperation,whichis:
the loading terminal such as the sampling and gas +foranLNGloadingtransferor
analysis. For CIF and CFR (and DES according to foranLNGunloadingtransfer
Incoterms2000)contractstheresponsibilityisreversed.
For simplicity, the parties may also make a commercial
Bothbuyerandsellerhavetherighttoverifytheaccuracy decisiontomutuallyagreeafixedgasquantity/volumeto
ofeachsystemthatisprovided,maintainedandoperated estimateEgas displaced and/orEgas to ER.
bytheotherparty.

The determination of the transferred energy usually
happens in the presence of one or more surveyors, the
ships cargo officer and a representative of the LNG
2.2. GENERAL SCHEME OF THE
terminal operator. A representative of the buyer can also MEASUREMENT OPERATIONS
bepresent.
The objective is to measure the quantity of energy

loaded from production facilities into an LNG carrier, or


Inallcases,thetransferredenergycanbecalculatedwith
unloaded from an LNG carrier to a receiving terminal.
thefollowingformula:
For shiptoship operations, the objective is to measure

the quantityof energytransferredfromone LNGcarrier


E V LNG D LNG GCV LNG E gas displaced
toanotherLNGcarrier.
Egas to ER , if applicable
From the above formula, it can be inferred that five
where: elementsmustbemeasuredand/orcalculated:

E = the total net energy transferred from the LNGvolume,
loadingfacilitiesto the LNGcarrier,from LNGdensity,
theLNGcarriertotheunloadingfacilities LNGgrosscalorificvalue,
or from one LNG carrier to another LNG energy of the gas displaced during the transfer of
carrier (shiptoship LNG transfer). In LNG,
international LNG trading, the energy energy of any gas consumed in the LNG carriers
transferred is most frequently expressed engineroomduring(un)loadingoperations.
in millions of British Thermal Units (106
BTUorMMBTU)althoughthisisnotaSI A graphic overview of the measurement scheme is
energy unit. Therefore, MMBTU is the shown in the figure Flowchart for determining the
preferred unit in this handbook. A energytransferred(seetheflowchartinSection2.2.5)
conversion factor table for other
commonly used energy units (such as 2.2.1. LNG Volume
MWh)canbefoundinEnclosure1.
Thestandardmethodchosenformeasuringthevolume
VLNG = thevolumeofLNGloadedorunloadedin of LNG transferred is based on the LNG carrier's
3 instruments, mainly the use of level gauges and
m .
calibrationtables.
7
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

For mostof thevessels,gauging hasbecome automated systemsabletoperformthesecorrectionsautomatically.


via the LNG carriers custody transfer measurement Itisrecommendedtousethemillimetreasthesmallest
system.Thesesystemsarecapableofdrawingupreports unitofdimension,whenapplyingatankgaugingtable.
of the volume of LNG on board at any time during
(un)loading.Thisisachievedbyconvertingthemeasured In some cases the LNG carrier must be completely
LNGlevelsineachcargotankintothecorrespondingLNG emptiedaftertheunloadingoperation,e.g.beforealong
volume in the cargo tank via the leveltovolume period of inactivity. In this case a special procedure
conversion tables and by applying correction factors for explained in Section 3.2.6 is followed for determination
trim, list and temperature and then by totalling the ofthevolumetransferred.
volumes in all the individual cargo tanks. Further details
aregiveninSection3.2.7. Before loading operations, the LNG carrier may be in
readytoload condition or otherwise, may require
Usually a quantity of LNG, called a 'heel', remains on gassingupand/orcoolingdownoperations.Inthiscase
board after unloading so as to keep the tanks cold. aspecialprocedureexplainedinSection2.6isfollowed
However,operatorsmaysometimesprefertostripoutthe fordeterminationoftheenergyandvolumetransferred.
cargotankspartiallyinordertomaximizetheLNGdelivery
or totally before the LNG vessel is scheduled for dry Unless parties explicitly agree otherwise (see below),
docking. gas flow is stopped and appropriate gas valve(s) to
engine room shut and sealed during and between the
openingandclosingcustodytransfersurveys.
Determinationofthevolumetransferredrequirestwosets
ofmeasurements,aninitialonebeforestartingloadingor The possibility of using LNG and/or boiloff gas as fuel
unloading and a final one at the end of the procedure. fortheshipduringtransferisconsideredinSections2.1,
Thesearecalledtheopeningandclosingcustodytransfer 2.3.4and12.2.
surveys(CTS)respectively.TwoLNGvolumesresultand
thedifference betweenthelargervolumeandthesmaller Since the calculation methods described in this
volumerepresentsthevolumeofliquidtransferred. handbook are based on volumes of LNG and LNG
vapour before and after transfer, any use of LNG,
Foranaccuratevolumemeasurementitisrecommended regasified LNG and/or LNG vapour during the transfer
that LNG piping on the LNG carriers deck including should be fully accounted for by correction of VLNG,
manifolds be in an identical inventory condition during according to the terms of the LNG purchase and sales
both custody transfer surveys (CTS). The piping should agreement.
either be completely filled with LNG both during the
openingcustodytransfer(i.e.before (un)loading)and the Note: In-line measurement of LNG quantity
closingcustodytransfer(i.e.after(un)loading)or,provided
that draining is possible before the closing CTS, Coriolis mass flow meters and ultrasonic flow meters are
alternatively be drained during both the opening and in use at some (un)loading terminals. However, at the
closing CTS. Where the piping is drained before or after time of writing, their use as part of a ship-shore custody
the CTS measurement, it should be done for sufficient transfer measurement system is not yet conventional.
timetofullyemptythepiping. This is mainly due to their high cost, the inability of these
flow meters to handle high flow rates and proving
As good practice it is recommended that the initial level issues. For small scale LNG transfer operations these
gauging should be made prior to any cooling down meters can be used as secondary (or even as primary
operation, i.e. after the (un)loading arms have been system), if agreed upon by the parties in their
connected but before any ships liquid and vapour commercial sales conditions, or just as (operational)
manifold valves have been opened. Where the opening verification for the parties involved. A further informative
CTS is conducted prior to commencement of tank cool discussion can be found in Appendix 1.
down, the CTS reading, where automated, may show
some liquid in the tank(s). The system should have the
capabilityofzeroingsuchreadingsforlevelandvolume, 2.2.2. LNG Density
sinceotherwiseanyliquidrecordedatcommencementwill The density of LNG is determined by calculation from
bedeductedfromthefinalCTSvolume. the measured composition of the LNG transferred and
the temperature of the LNG measured in the LNG
Thefinallevelgaugingreadingshallbemadeassoonas carrier'stanks.
possible after completion of (un)loading with liquid and
vapour arms (or flexible hoses) drained and inerted, and
withliquidandvapourmanifoldvalvesclosed. 2.2.3. LNG Gross calorific value
The composition of the LNG is used to calculate the
The level gauge readings shall be determined by the grosscalorificvalue.
arithmetic average of several successive readings at
regular intervals. Further details are provided in Sections
3.2.1,3.2.2and3.2.3. 2.2.4. Energy of the gas displaced by the transfer
of LNG
Intheeventoffailureoftheprimarylevelgaugingdevice,
anauxiliarydeviceshouldbeused. This energy is calculated according to the composition
andvolumeofthegasdisplaced,andthepressureand
Level corrections are to be made using correction tables temperature of the gas inside the tanks of the LNG
providedfortheLNGcarrierastankgaugetablesfortrim, carrierbeforeloadingorafterunloading.Thecalculation
list and also for temperature. Most LNG carriers are procedureisexplainedinSection12.1.
equipped with process control systems or standalone
8
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

2.2.5. Flowchart for determining the energy transferred

2.3. INSTRUMENTS USED



2.3.1. For the determination of the LNG volume Note: Automated systems
ForthedeterminationoftheLNGvolumethefollowing The calculation to determine LNG volume may be
arerequired: automated by processing the level, temperature and
the LNGcarrierscalibrationtables,includingthe pressure measurements, taking into account the
main gauge tables for each tank and different above mentioned calibration and correction tables to
correction tables accounting for list and trim produce a report meeting CTS requirements. LNG
variances (if any) for the main and secondary carriers may be fitted with certified custody transfer
gauging systems, tank contraction tables for measurement systems for this purpose. See Section
Mosstype and SPBtype cargo containment 3.2.7.
systems, and possibly, other correction tables
accordingtothetypeoflevelmeasuringdevices,
2.3.2. For the determination of LNG density
the equipmentformeasuring thelevelof LNGin and gross calorific value
the LNG carrier'stanks.Each cargo tank usually
has two level gauge systems installed, one The determination of the density and the gross
designatedas'main'or'primary'andtheotheras calorificvalueoftheLNGtransferredismadeonthe
'secondary'. Capacitance and microwave (radar) basis of the average composition of the LNG
level gauging systems are widely used as obtainedby:
primary CTS systems onboard LNG tankers, continuousordiscontinuoussamplingofLNGin
backedupbyasecondaryCTSsystemgenerally the LNG transfer line(s) between the ship and
consistingofafloatgaugingsystem, theterminal,
recently a laser system (called LIDAR) was gaschromatographicanalysis,
introduced in the industry but at the time of followedby:
writing its use is restricted to very few LNG
carriers. A further informative discussion can be acalculationbasedontheaveragecomposition
foundinAppendix2.SomeveryoldLNGcarriers of the LNG, its average temperature and the
still in operation are fitted with nitrogen bubbler coefficients given by the National Bureau of
systems, which rely on good knowledge of the Standardsforthedensity[9],
LNGdensitytogiveaccuratereadings, acalculationbasedontheaveragecomposition
of LNG and characteristics of elementary
temperatureprobesdistributedovertheheightof components (GCV, molar volume, molar
theLNGcarrier'stanks, weight) given by reference tables or standards
othermeasuringdevicesrequiredforapplyingthe forthegrosscalorificvalue.
correctionfactors.

9
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

AtaloadingterminalLNGsamplingand analysisare procedures accepted by both parties to accurately


made in the LNG transfer line(s) prior to possible measurethegasenergyconsumptionintheengine
flashing (vaporization) in the ships cargo tanks. If room between the opening and closing custody
flashing occurs in the ships cargo tanks, then this transfersurveys(CTSs),andthatthisonboardgas
causesaminorchangeinLNGcompositionsincethe energy consumption is taken into account as Gas
most volatile components (typically nitrogen and to Engine Room as per the general formula in
methane)arepreferentiallyvaporizedandreturnedto Section2.1.However,forsimplicitythepartiesmay
shoreviathevapourreturnline.Therefore,thiseffect makeacommercialdecisiontomutuallyagreetoa
shouldbeavoidedifpossible,orotherwiseminimized, fixedgasquantity/volume.
e.g. by ensuring that the tank pressure in the ships
cargo tanks is sufficiently higher than the saturated
vapourpressureoftheLNGbeingloaded. 2.3.5. Periodic instruments recalibration

Notes for information only: Itisrecommendedthat,unlessitisspecifiedbythe


fiscal authorities or by the Classification Society,
A novel LNG analysis method Buyer and Seller agree on the periodicity of
At the time of publication of this fourth edition, a different recalibrationintervals,e.g.ateachdrydocking.
measurement device for analysing LNG composition based
on direct analysis in the LNG transfer line(s) hence
eliminating the need of an LNG sampling device, vaporizer
and gas analyzer, is being tested in a few pilot applications
(see Section 7.6 Raman spectroscopy).
2.4. STANDARDIZATION
International standards exist for the classical
methods and techniques used for LNG Custody
2.3.3. For the determination of the energy of Transfer such as ISO 6976 for calculation of the
displaced gas GCVofgas.
The energy of the displaced gas can be determined
from: On the other hand, a number of existing LNG
supply, shipping and purchase agreements specify
samplingofthegasdisplaced, GPA22612000forgaschromatographyandHM21
or GPA 21452009 and GPA 21722009 for the
a gas chromatographic analysis of this sample
calculationoftheGCVofthegas.BuyerandSeller
gas,enablingtheGCVtobecalculated,
mayapproveoneoftheseeditions,usuallythemost
pressure and temperature measurements within recentedition.
theLNGcarrier'stanks.
As far as methods and techniques dealing with
However, for the determination of the energy static measurement procedures for LNG are
displaced, some parameters such as pressure, gas concerned, it should be noted that ISO has issued
composition and temperature can be estimated from numerous international standards, (see Enclosure
experience and taken as constant for both custody 2). The recommendations included in these
transfersurveysbeforeandafter(un)loading. documents and future international standards are
appropriatetobeconsideredfornewagreements.
For instance, the displaced gas may be assumed to
be a fixed mixture of nitrogen and methane, or pure
methane. This assumption will hardly increase the
overalluncertainty. 2.5. PARTIAL LOADING OR

UNLOADING OF LNG
2.3.4. For the determination of the energy of CARRIERS
Gas to engine room
Inrecentyearstherehasbeenaworldwideincrease
Partiesmayexplicitlyagreetoallowgasconsumption inshorttermandspotcargoLNGtrading,involving
in the LNG carriers engine room (also including the twonewoperatingtrendsinLNGshipping:
gasburntbytheshipsGCU)duringthetimebetween
more and more LNG shippers are using LNG
the opening and closing custody transfer surveys
carriersasfloatingLNGstorage,
(CTSs). This could be to ensure low airemission
operation in the engine room whilst at berth and so several LNG shippers have considered and
may favour the use of boiloff gas perhaps some havecarried out partial unloading and/or
complemented by regasified LNG rather than fuel oil partialloadingofoneorseveralcargotanksof
intheengineroom.ThispracticemayenabletheLNG LNGcarriers.
carrier operator to comply with MARPOL Annex VI
(revisionOctober2008). When performing such operations, due attention
For (re)loading operations or shiptoship LNG shouldbegivento:
transfer operations, this could also be done on the
LNGcarrier(s)inordertohandletheboiloffgas(flash safe ship/shore operating practices and
gas) produced during such operations, hence procedures,
reducingthevapourreturnedtoshoreortothevessel proper LNG ship/shore custody transfer
beingunloaded. procedures.

It is recommended in this case that the LNG carrier Please refer to Appendix 3 for recommended safe
has proper measuring equipment on board and practicesforpartial(un)loading.

10
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

2.6. Gassing-up and cooling down 2850MWh for a complete gassingup operation.
Thisgassingupoperationwilltakeapprox.20hours
operations to complete depending upon the LNG supply and
2.6.1. Gassing-up operations thevaporizationrateonthevessel.Thegassingup
tables should be reviewed and agreed upon by all
When an LNG vessel is delivered or after dry dock, parties at the preliminary meeting prior to
the cargo tanks are often filled with inert (exhaust) commencinganyoperations.
gas. As inert exhaust gas contains carbon dioxide
which will freeze during loading, it must be replaced There are some terminals which correct the
withwarmLNGvaporpriortocoolingdownthetanks theoretical volume taking into account the
in preparation of loading. This process is called measurement conditions when these are different
gassingupandisnormallyprecededbyaprocessof thantheonesinthecertifiedgassinguptables.
dryingandinerting.
Other alternatives to measure the energy for
Drying can be done with hot air (or nitrogen) and gassingup(especiallyforoldLNGvesselswhichdo
inertingisperformedtoremovetheoxygenoutofthe not have certified gassingup tables) are the
cargotanksandreplacetheair/oxygenbyinertgases followingones:
(exhaustgasesoftheshipornitrogen).Incasedrying
and inerting are performed with nitrogen, both steps a) the procedure stated in the LNG vessel
canbecombinedinoneoperation.Thereasonbehind OperationsManual,
thepreliminaryoperationsofdryingandinertingisnot b) to apply twice the theoretical volume of
to directly replace air by natural gas due to safety each cargo tank to obtain the theoretical
reasons (hence avoiding an explosive atmosphere energyforgassingup,
duringtheoperation).
c) to measure the delivered LNG (in cubic
Onceinerted,theshorewillsupplyLNGwhichissent meters)duringthegassingupoperationby
to the main vaporizer of the ship to produce vapor meansofanLNGmassorflowmeter,
warmer than the dew point temperature in the cargo
d) to measure the difference in level in the
tank. This vapor is then injected at the top of the
shore LNG tanks. In this case, it is
cargo tank to displace the inert gas. The increase of
necessary to take into account any other
pressureandthedifferenceindensityforcestheinert
filling (e.g. cold circulation) or emptying
gases out of the cargo tanks. The exhaust gas is
(e.g. send out for regasification purposes)
generallydirectedtotheships(forward)ventmast,to
operationsofthetanks.
aventonshoreorburntintheterminalsflare(atthe

beginning it is fully inert and gradually becomes a
2.6.2. Cooling down operations
mixtureofnitrogenandinertgases(whichpercentage
is decreasing) and natural gas (which percentage is Cooling down operations are performed to slowly
increasing). This process continues until the exhaust reduce the temperature of the cargo tanks close to
gasismeasuredtohaveapprox.9899%ofmethane that of the LNG to be loaded in order to avoid any
and the CO2content is less than a certain low structural damages by thermal shock or stress to
threshold(e.g. 0.1%).Oncethisisaccomplished,the the tank construction. The target is to reach a
vessel is ready to receive cold vapor and start the certain reference temperature according to the
coolingdownprocess. performancecriteriastatedintheoperationsmanual
orcooldowntablesoftheLNGvessel.
The quantity of LNG required for this operation
depends onthesizeandconstruction of the vessels Cooling down operations are generally performed
cargo tanks. The gas quality considered for gassing onanLNGshipofwhichthecargotanksareunder
up is usually the same as the quality for cooling and (warm)naturalgas(e.g.afteralongballastvoyage
for loading, but some loading/reloading facilities can duringwhichthecargotankshavebeenwarmedup
use different gas qualities for each operation. The or (immediately) after gassingup operations).
energy required for gassingup the cargo tanks is However it is also possible to cool down an LNG
statedintheterminalrulesorinthecontractbetween ship ofwhichthecargotankshavebeenfullydried
theparties,butthemostcommonlyusedtechniqueis and inerted with nitrogen, hence avoiding gassing
to use the certified gassingup tables of the LNG upoperations.
vessel.
LNG is supplied by the LNG terminal and it is
The certified gassingup tables apply a volume of vaporized and injected at a controlled rate into the
natural gas which is between 1.7 and 2 times (each shipscargotanksthroughspraynozzles atthetop
LNG vessel has a coefficient) the volume of each of the cargo tanks to avoid thermal shock in the
cargo tank, which is the theoretical volume of gas to tanks and its devices (such as pumps, pump
be supplied to each cargo tank for the purpose of columns, probes). Once the LNG is vaporized,
gassingupthiscargotank. there is an increase of pressure in the cargo tanks
and the gas return is sent to the terminals vapour
Each vessel will have a gassingup table for each return system or to the terminals flare to keep the
tank on board. These are provided by the tank pressureinthecargotanksundercontrol.TheLNG
manufacturer and confirmed by an independent vessel can also help reducing its cargo tank
surveyor during construction at the shipyard. Each pressurebyburninggasasfuel(intheenginesorin
tableisdesignedspecificallyforthatparticulartypeof thegascombustionunit).
containmentsystem.Forexample,a145000msize
vessel will require approx. 420 m3 of LNG and

11
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

Therequiredtemperatureneededbeforeloadingcan the range between 40C and 130C. In this


be started is identified in the Operations Manual or casetheoperationcantakeabout6to8hours.
the cool down tables provided by the tank
manufacturer.Eachcooldowntableisspecifictothat b) MosstypeLNGvessels
tank type, is issued by the manufacturer during
construction and is verified by an independent In this type of vessel the temperature reference is
surveyorpriortothevesselbeingdelivered.Thecool the equatorial temperature of the cargo tanks. The
down tables identify the quantity and length of time referencetemperatureforloadingaMosstypeLNG
required to complete the operation prior to loading. vessel normally spans the range between 110C
Thegasqualityconsideredforcoolingdownisusually and130Cdependingontheoperationsmanualof
thesamethanthequalityforloading(andgassingup theLNGvessel.
if there is any), but some loading/reloading facilities
can consider different gas qualities for each Inthiscase,thecertifiedcoolingdowntablescould
operation. The energy required for the cooling down give:
operationsofthecargotanksisstatedintheterminal
rules or in the contract between the parties, but it is theLNGvolumeneededtoreducetheequator
usualtoapplytheuseofthecooldowntablesorone temperatureofthetankbyonedegreeCelsius.
ofthefollowingalternatives. The total LNG quantity required for cooling
down the cargo tanks will be calculated by
Coolingdowntables multiplying the difference between the initial
The most common method is the use of the cool equatortemperatureandtemperaturereference
downtables.Basedonthesizeofthecargotank,the bythevalueofcubicmetersthatareneededto
manufacturer calculates how many cubic meters of lower the equator temperature by one degree
LNGorwhatenergycontentarerequiredtolowerthe Celsius.
temperatureinsidethetankonedegreeCelsius.Then the LNG volume and energy needed to reach
basedonthevaportemperatureinsidethetankatthe the reference temperature (as the equatorial
beginning of the operation, the quantity is calculated temperature).
accordingly. The time it takes to complete a cooling
down operation depends on the temperature prior to Nozzlepressure
startingtheoperation.Thecoolingdowntablesshould
be reviewed and agreed upon by all parties at the The cooling down operations are performed by
preliminary meeting prior to commencing any injectingLNGthroughspecialspraynozzlesintothe
operation. cargo tanks. The flow is fully dependent on the
applied pressure. The number of nozzles, the
Cooling down operations are generally faster than average pressure and the duration of the spraying
gassingupoperations.Ittakesapprox.10to12hours foreachtankcanbeusedtodeterminethevolume
for a membrane type LNG vessel and approx. 20 ofLNGusedforthecoolingdownoperations.
hoursforaMosstypevessel.

a) MembranetypeLNGvessels Incase the cooling down operations are performed
onanLNGshipofwhichthecargotanksareunder
Inthistypeofvesselthereferencetemperatureisthe nitrogen,thecooldowntablesmaybeusedaswell,
average temperature (vapor phase) of the pump howeverthisshouldbeagreeduponbyallparties.
tower in each tank excluding the first top or two top
sensors (depending on terminal rules/contract
betweentheparties).
3. VOLUME MEASUREMENT
The reference temperature for loading a membrane
typeLNGvesselshallbeapprox.130C.
3.1. GAUGE TABLES
3.1.1. Use of gauge tables
It is usual to use the certified cooling down tables.
These tables give the energy required to cool down Thegaugetablesarenumericaltableswhichrelate
eachtankfromitsarrivaltemperatureto130C. the height of the liquid in an LNG carrier's tank to
thevolumecontainedinthattank.Thevolumemay
Thecertifiedcoolingdowntableshavebeenmadeup need to be corrected taking into consideration
supposing a certain gas type in order to obtain the variousfactors.
energy. Once cooling down energy has been
obtained,itispossibleto obtain anLNGcubic meter An independent surveyor usually produces the
equivalentforthiscoolingdownoperation. gaugetablesduringthebuildingoftheLNGcarrier.
Theytakeintoaccounttheconfigurationofthetank,
Thetablesaredividedintotwosections: its contraction according to the temperature of the
liquidandthevolumesoccupiedbyvariousdevices,
Warm conditions. These are used in case the e.g.cargopumps.
averagetemperatureinthecargotanksishigher
than 40C. The table normally spans the range Thecalibrationtablesareusuallydividedinto:
between +40C and 130C. In this case the
operationcantakeabout10to12hours. maingaugetables:height/volumecorrelationin
idealconditions,
Cold conditions. These are used in case the
average temperature in the cargo tanks is equal
toorlowerthan40C.Thetablenormallyspans

12
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

correctiontablestakingintoaccountactual thelistoftheLNGcarrier(seeFigure1,),
conditionsoftheLNGcarrieranditsmeasuring
instruments. thetrimoftheLNGcarrier(seeFigure2,),

For each LNG carrier there is one main gauge table These corrections can be positive or negative. So
pertank.Generallythevolumesaregivenforheights therealheightwillbeequaltothealgebraicsumof
varyingcmbycm,thevolumeforintermediateheights the height reading, the correction for list and the
in mm being calculated by interpolation. An example correction for trim. These tables are made up in
ofagaugetableisgiveninAppendix4(tableA41). degrees for the list and in metres for the trim, with
fixedstepsofvariation.Forintermediatevalues,the
To avoidtheseinterpolations,whichcanbeasource correctionwillbecalculatedbyinterpolation.
of inaccuracy, the mostused parts of the gauge
tables i.e. heights between 10 and 60 cm and In practice, the LNG carrier's cargo officer will
heightscorrespondingtoavolumebetween95%and usuallymanagethevessel'sballasttoobtainzeroor
98% ofthe totalvolume of the tank aresometimes alimitedlistandtrim.
developedinafinegaugetableandthevolumescan
be calculated mm by mm. This then reduces the 3.1.2.2 Corrections according to the tempera-
determination of the volume to a mere reading in a tures in the liquid and gaseous phases
table(seeAppendix4,tablesA42andA43). Thecorrectionsarerelatedtothevolumevariations
resultingfromthecontraction ofthetanksandtheir
The examples used in this section are taken from a insulationaccordingtothetemperatureoftheliquid
vessel with prismatic cargo tanks. The same andgaseousphases.
principles generally apply to those vessels with
sphericalorothershapesofcargotanks. ThisphenomenonissignificantforLNGcarrierswith
selfsupportingtanks.Appendix4givesanexample
Various methods exist for establishing the gauge ofthesetables(seetableA44).
tables.Themainmethodsare:

macrometrologywithtapes, 3.1.3. Approval by authorities
alasermeasuringsystem, The gauge tables may be approved by either the
aphotogrammetricmeasuringsystem. authoritiesofthecountriesconcernedwiththeLNG
sale and purchase or by independent sworn
For details of calibration procedures for tanks, measurers. In practice, largely due to the
referencecanbemadetoexistingISOstandards(see importance for LNG shipping of the Japanese
Enclosure2). market and its requirement for NKKK certification,
the great majority of LNG carriers have such
certification.
3.1.2. Correction tables
This approval may be valid for a limited duration,
The gauge tables are completed with correction generally10to12years,orlessdependingonLNG
tablesestablishedaccordingto: terminal requirements, provided there are no
modificationstothetanks.FortheEuropeanUnion,
theconditionoftheLNGcarrier(trim/list),
thisapprovalcorrespondstoacommunitydirective.
the average LNG temperature in the tank that
influencescontractionorexpansionofthetank, When an LNG carrier is put into operation, a list of
all works on the tanks must be supplied, and the
thetemperatureinthegaseousphase,and/orthe tanksmustbeinspectedforanymodificationswhich
density of the LNG influencing the level mightaffectthevolume.
measuringdevices.
Tank gauge tables may also provide an example of In the case of any distortion or modification to a
how to conduct the volume calculation using the tank,thegaugetablemustbeadjustedaccordingly.
measurementsprovided.
Theapplicableandconfirmedstandardscorrespond
It should be noted that LNG carriers normally have to:
two level measurement devices in each cargo tank
(and often of two different types). Correction tables ISO 8311:2013 Refrigerated light hydrocarbon
are specific to a level gauge. Using the correction fluids Calibration of membrane tanks and
tables for the wrong gauge can result in significant independentprismatictanksinshipsPhysical
inaccuracies. measurement,

3.1.2.1 Correction according to the condition of
the LNG carrier
The gauge tables are established for an LNG carrier
withzerolistandtrim.Therefore,itwillbenecessary
tocorrecttheheightreadingtotakeintoaccountalist
oratrimwhichisnotzero.

Correctiontablesaremadeupaccordingto:
thepositionofthegaugeinthetanks,

13
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

FIGURE 1



Listrepresentedbytheangleindegreestoport.Inthisillustrativecase,thecorrectionwillbenegative.


FIGURE 2



Trimexpressedinmetresorfractionsofametre,accordingtothedifferenceinbowandsterndrafts.

Thecorrectionisnegativeorpositivedepending onwhetherthe bowoftheLNGcarrierisdeeperinthewateror


otherwise.Inthisillustrativecasethecorrectionwillbenegative.

14
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

3.2. INSTRUMENTS AND METHODS placedatregularlyspacedintervals alongthewhole


length of the tubes. The resulting assembly forms a
FOR MEASURING THE LEVEL seriesofcylindricalcapacitors,havingthesametotal
OF LIQUID IN THE LNG heightasthecargotankoftheLNGcarrier.
CARRIER'S TANKS
The LNG, according to its level, will fill the space
between the concentric tubes. The liquid affects the
3.2.1. Main liquid level gauging devices dielectriccharacteristics of the capacitors formed by
the tubes such that, by measuring the change in
Themaintypesofgaugesare:
capacitance, the height of the LNG in the annular
electricalcapacitancetypegauge, space, and hence the level in the tank, can be
determined.Thecontractionofthetubeassemblyat
floattypegauge, lowtemperaturemaybetakenintoaccounttocorrect
radartypegauge, thelevelmeasurement.

lasertypegauge:arecentlydevelopedsystem. Theaccuracyofthemeasurementresultingfromthe
calibration of the dimensions andthelinearityofthe
Currently ISO 181321:2011 and ISO 181322:2008 capacitor and of the electronics should be, for the
are valid for the abovementioned level gauging gaugeasawhole,5.0mm[1].
devices.
3.2.1.2 Float type level gauge
Any of these gauging systems can serve as main Measurements are made with a float hanging on a
instrument for liquid level measurement inside an tape or a ribbon (see Figure 4). According to the
LNG cargo tank. Usually (but not always) two or leveloftheliquid,thefloatisdisplaced,andthetape
sometimeseventhreetypesoftheabovementioned or the ribbon on which it hangs is unrolled or rolled
systemsareinstalledineach cargotankoftheLNG up on a drum whose rotation is recorded. This
carrier. One of these should be agreed as the main enablesthepositionoftheprobe,andthusthelevel
(primary) level gauging device by buyer and seller. ofliquidinthetank,tobeknown.
Thestandbyorbackupgaugingsystem(s)is(are)to
be considered as the auxiliary (secondary) gauge Withfloatgauges,itisnecessarytotakeintoaccount
systems. the shrinkage of the ribbon, according to the
temperatureofthegaseousphaseandtheheightof
A few LNG shipping agreements do not specify the the liquid, and the density of the LNG, which will
type and only specify the required accuracy (e.g. influence the float buoyancy. The correction tables
5.0 mm or better). On old ships other systems may willtabulatethecorrectionsfortheseeffects.
befound,suchasnitrogenbubblingdevices,butthe
accuracy of these is generally lower and requires a The corrections for temperature are required only in
good knowledge of the actual LNG density in the thecaseofastainlesssteelribbon.Inthecaseofan
cargotank. invar ribbon, the shrinkage is much less and is
generallyconsideredasnegligible.
Accuracymay be verified at the time of gauging, as
follows: The accuracy of this type of gauge, designed for
- forelectricalcapacitancetypegauges,by marine application, is in the range of 4 mm to
anonlinevalidationsystem, 8mm.

- forradartypegauges,dependingonthe 3.2.1.3 Radar (microwave) type level gauge
design,eitherbyaverificationpinorby
comparisonwiththedeterminedlengthof The radar or microwave type gauge works on the
eachstillpipesegment, same principle as a ship's radar(seeFIGURE 5).A
transmitter/receiver is mounted on the top of the
- forfloattypegauges,eitherbycomparison cargo tank and emits microwaves vertically down
withtheothergaugingsystemorthe towards the surface of the liquid. The signal is
stowed/groundedinstrumentslevel reflected from the surface, received by the
readings. transmitter's antenna and sent back to the control
panel.Thesignalisthenprocessedtodeterminethe
For the measurement of liquid levels in terminal distanceoftheliquidsurfacefromthetransmitterand
berths exposed to the open sea, filtering systems hence ullage. One or more certified transmitters/
approved and accepted by sworn surveyors as receivers are usually carried as spares in the event
suitable to be integrated in CTS operations may be ofafailureofthisequipment.
considered.Thesefilteringsystemswerespecifically
developed for offshore use where large level Since all the level detection components are
fluctuations may be experienced. ISO 10976:2012 mounted external to the cargo tank, the microwave
includesinformationonthesesystems. system allows for the possibility of changing the
gaugeinservice.
3.2.1.1 Electrical capacitance type level gauge
Theradargaugerequiresawaveguidewhichislike
The electrical capacitance gauge (see Figure 3) a stilling well. However, this well is a critical and
consists of two concentrictubes made of aluminium complexcomponentresemblingagunbarrel.
or stainless steel, depending on the construction of
the LNG cargo tank.Theinnertube is supportedby
the outer tube by means of concentric insulators

15
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

There are several types of gauging pipes used for appointarepresentative.Aswornsurveyormayalso


microwave type CTS level gauges. Particular bejointlynominatedbybothpartiestostandasthird
attentionshouldbepaidtobottomattenuators,which and neutral party, and witness and record the CTS
can limit minimum gauging height and therefore readings.
accuracyduetotheirmountings.
The readings of the levels of liquid in the tanks are
Theaccuracyofthistypeof gaugecan be5.0mm takenafterthereadingsofthelist(portorstarboard),
orbetter. and the trim (bow or stern) of the LNG carrier. All
thesereadingswillbetakeninto accountin order to
3.2.1.4 Laser type level gauge determine the corrected liquid level and hence the
The laser based sensor technology (LIDAR) is a LNG volume in each of the cargo tanks. The
recentlyintroducedgaugetype(seeAppendix2). temperatures of the liquid and the gaseous phase
are also measured (see Section 4.1 and Section
The laser unit is installed above the deck, isolated 4.2).TheabsolutepressureineachLNGcargotank
fromthetankbyasightglassandtargetingtheliquid ismeasuredaswell.Ifthepressuremeasurementis
level surface with a low power laser beam. This gaugeinsteadofabsolute,theatmosphericpressure
beam is protected from adverse influences on its isalsoread.
measurements arising from within the tank by an
ordinarystillingwell.Thestillingwellusedonalaser Parties should also agree in advance what actions
based system is not used as a wave guide, its aretobetakenintheeventofafailureofanypartof
construction is simple and it remains relatively the volume measurement system (e.g. consider
inexpensive. secondarylevelgauging,manuallycalculatevolumes
withtanktables,etc.).
The time interval between the transmitted and
received pulses is accurately measured and 3.2.3.1 Reading of the level with float gauges
processed to determine the distance of the liquid Itisgoodpracticethatallfloatgaugesareloweredto
surface from the transmitter and hence the ullage theliquidlevelwellbeforeanyreadingistaken.This
height. should allow for sufficient time to cool down either
thegaugeribbonorwireandtoensureshrinkage,if
The level measurement remains unaffected by any,isstabilized.Itisalsogoodpracticetoverifythat
changes in the LNG composition or by the changes the calibration seals on the float gauge unit are
in physical properties of the still pipe. When intact.
performinglevelmeasurements atthe bottom ofthe
tank, echoes or reflections near the bottom do not Float gauges normallyhavetworeferencereadings,
affect the accuracy of the system. The gauge will upper and lower. The upper reading can always be
alsoregisteradynamicchangeintheLNGlevel.The checked. Depending on the manufacturer, this
accuracyofthistypeofgaugeis7.5mmorbetter. correspondstoalockedpositionofthefloatgaugein
its stowedcondition oraresting position ofthefloat
on the valves isolation gate. These reference
3.2.2. Timing of the level measurement readings are reported on the ships CTS calibration
certificateandtheyarealsostampedonaplatefixed
GeneralguidelinesaredescribedinthestandardISO to the gauge itself. It is advisable that the upper
6578 Refrigerated hydrocarbons liquid static reference is checked before the cargo tank liquid
measurementcalculationprocedure. levelsaretaken.

Itisimportantthatthelevelreadingsarerecordedin The cargo liquid level is never still and a slight
astaticandclosedconditionoftheLNGcarrier,with movement of the float is always to be expected.
noflowofeithergasorliquid. Even when the ship is not subject to any wave
motion,thereadingswillalwaysbeanaveragelevel
Asfaraspracticable,theperiodbetweenthetimeof betweenupperandlowerpeaksoflevelreadings.
measurement and start of cargo transfer should be
narrowed to the minimum achievable. The same If no fluctuation of the liquid level is detected, this
shouldapplyforthefinalcargocustodytransferlevel may indicate a possible sticking of the float inside
measurement,oncetimeisallowedtocorrectlydrain thegaugewell.Ifthisisthecase,itisrecommended
cargo lines (as appropriate) and stabilize the liquid that the float is moved up and down a number of
levels. The same temperature and inventory timesinordertofreeitandobtainreliablereadings.
condition of the ships LNG cargo lines should be
maintained duringbothopeningandclosingcustody On occasions when the LNG carriers tanks are
transfer surveys. Wherever it is intended that the completely empty, the lower reference of the float
vessel heels out (is completely emptied) and cargo can alsobe checkedforcorrectness.This isusually
lines are empty upon completion of discharge, it is calledthefloatsgroundedposition.
necessarytoensurethatthecargolinesaredrained
forsufficienttimepriortotheclosingCTS,toensure Wheneverremotefloatreadingsare availableinthe
stabilization of the temperature of the lowest ships control room, itshould be checked that these
temperaturesensor. are approved for commercial transactions before
they can be used for CTS purposes. Float gauges
3.2.3. Readings arealwaysstowedintheirfixedupperpositionwhen
Itisgoodpracticethatallreadingsarewitnessedby sailingsothattheribbonorthetapedoesnotbreak
both parties, buyer and seller. Each party should duetoliquidmovements.

16
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

significantly alter wave velocity and should be


3.2.3.2 Reading of the level with capacitance considered.
and radar gauges
The methods of reading are normally agreed Oneavailabletypeofmicrowavegaugeusesseveral
between the seller and buyer. Usually the observed defined targets made of PTFE (Teflon ) at known
cargotanklevelisaveragedover5readingstakenat positions along the stilling pipe. This enables the
specifiedandregularintervals. system to precisely calculate the microwave
propagation speed and adapt it to locate the liquid
The above method may not be suitable for LNG level. The PTFE target position is usually known at
terminalberthsexposedtotheopenseabecauseof +20C. In order to determine the wave propagation
thehighfluctuationsoftheliquidlevels.Tocopewith speed, the position of the targets is corrected
this, some custody transfer measurement systems according to the shrinkage of the still pipe, which is
manufacturers have lately developed filtering calculated as a function of the vapour phase
systems in order to average the level fluctuations temperatureabovetheliquidlevel.
and provide a realistic liquid level (see Section
3.2.1). Astocorrectionofreadingsforlasergauges,please
refertoAppendix2.
Level readings are normally available in the ships
cargo control room and sometimes in an Moderncomputerbasedsystemsusuallycanaccept
instrumentation room located in the accommodation trim and list data either manually or from external
block. sensors and automatically apply the corrections.
However, it is important to compare the observed
SomemodernonboardCTMsystemsusecomputer trim/list to the data from the trim/list independent
to process all the information, including averaging sensor.
the (5) level readings over time, temperature and
pressure, and draw on digital gauging tables to Depending on the sensors location, either sagging
produce a printed document containing all the ship or hogging of the LNG carrier may seriously affect
generated information required for the custody theaccuracyofthereadings.
transfer. However these seldom include gas
displacedorgastoengineroom,ifapplicable.

3.2.4. Correction of readings


3.2.4.1 Float gauge
Thereadingsmadeonthemeasurement appliances
shouldbecorrectedaccordingto:
list,
trim,
densityofLNG,affectingfloatbuoyancy,
coefficientofcontractionofthematerialandthe
insulationofthetanks;thiscoefficientismore
relevantinthecaseofselfsupportingtanks(see
Appendix4,tableA44),
temperatureofthegaseousphaseiftheribbon
orcableisnotmadeofinvar.
standardandfixedcorrectionsifapplicable.

The corrections are made using specific tables for


eachoftheabovecorrections.

3.2.4.2 Capacitance, radar and laser gauges


Inthiscase,onlythecorrectionsforlist,trimandthe
contraction of the tanks are normally taken into
consideration. For accurate level measurement the
contraction of the capacitance gauge at low
temperaturemayalsoneedtobeconsidered.

Two main types of radar (microwave) gauges are


currently available. One type is based on the
principle of estimated velocity of the microwaves
insideanatmosphereofpredeterminedhydrocarbon
composition. Within the worldwide range of
commerciallyavailable LNG compositions this has
little impact on the overall accuracy of the gauge
exceptfornitrogencontentinthevapourwhichmay

17
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

FIGURE 3: ELECTRICAL CAPACITANCE TYPE LEVEL GAUGE

1. Outertube
2. Innertube
3. Concentricelectricalinsulator
4. Isolationofinnertubesectionsbyagapordielectricplug
5. Isolationfromthetankbottom
6. Connectionsbetweenthesectionsoftheoutertubetomakeasingleelectricalconductor.
7. Datasignalconnectionsfromthegaugetothejunctionboxoutsidethecargotank
8. LNGcargotank

18
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

FIGURE 4: FLOAT TYPE LEVEL GAUGE


19
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

FIGURE 5: RADAR (MICROWAVE) TYPE LEVEL GAUGE


20
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

3.2.5. Use of spare level gauge liquid level is always within the certified minimum
gauging height. e.g. using onboard CTMS or cargo
When the main or primary level gauge cannot be tanktablesusingappropriatecorrectiontables.Inthe
used, the spare level gauge, also called the event the liquid level falls outside the certified
secondarylevelgauge,isusedtomeasurethelevel minimum gauging height, buyer and seller should
of LNG. Therefore it should always be in operation. agreeontheimmeasurablecargoquantity.
Usually, the secondary level gauge will also be The energy of this remaining LNG transferred either
calibrated and certified. However, if the calibration in the liquid or in gaseous form can also be
tables of the LNG carrier are available only for the determined by mutual agreement by the parties
main level gauge, a conversion table is required in takingintoaccountanytechnicallimitationstheymay
ordertotakeintoaccounttherespectivelocationsof have.
main and secondary gauges, or the statistical
differences between the two level gauge Note
measurements, and to evaluate the corresponding Certain parties have raised concerns over the
corrections which must be applied to level possibility of having a quantity of LNG
measurementbeforeusingthecalibrationtables. unaccounted for during closing of CTS under a
positive trim as a result of LNG being trapped in
Since it is possible that the primary level gauging the bottom part of uncovered corrugations of
systemcouldfailduringthecargotransferoperation, MARKI/MARKIIImembranesystems,wheresmall
and since it is strongly recommended to use the liquid wedges may form between the tanks
same reference system at both opening and closing bottom plate and the transverse corrugation in
CTS measurements to determine the cargo volume each of the exposed tanks bottom cells. The
transferred,itisthereforeadvisablethatbothprimary validity and extent of such liquid entrapment is
and auxiliary level gauging systems readings are currently unknown. This will require further and
recorded at the opening CTS measurement, before full investigation by independent and expert
the start of cargo transfer operations. It should also parties, before any conclusion can be made on
beensuredthatthesparelevelgaugeismaintained the subject. The conclusion of such investigation
withinstatedaccuracyatalltimes. should then be fully endorsed and certified by
independentandrecognizedswornsurveyors.

3.2.6. Complete unloading (tank stripping)
Cargotankstrippingwithstrippingpumps: 3.2.7. Automated systems
The LNG carrier may be fitted with an automatic
the LNG carrier needs to be able to continue system for the calculation of LNG and gas volumes
dischargingoncethecargopumpscannotlonger in each tank. The use of such a system, commonly
be kept running because the level in the cargo referred to as the ships custody transfer
tankistoolow, measurement system (CTMS) will facilitate the
process of determining quantities transferred during
these stripping operations require several loadingandunloading.
additional hours at the port. Stripping to lowest
levels is generally conducted prior to scheduled The CTMS processes data from tank level,
drydockingorforspotcargooperations, temperature, pressure sensors, etc. in realtime,
the remaining quantity after stripping can be taking into account the required corrections and
vaporised by warming up to a temperature at certified gauge table, to produce a calculation of
whichtheLNGcarrierisconsideredtobeempty volumes before, during and after transfer. By taking
of LNG. e.g. 80C (no more liquid ethane) or measurements frequently, data can be averaged to
40C (no more liquid propane). The improverepeatabilityofthecalculation.
vaporization of residual LNG is typically
conducted not at berth but at sea, after the Modern custody transfer measurement systems
unloadingisfinishedandthevesselhasleftthe typically comprise two discrete parts: (a) the tank
berth. gauging system providing corrected tank levels,
temperatures and pressures and (b) workstation(s)
and peripherals, usually located in the ships cargo
Forthecompleteornearlycompleteunloadingofan control room, to perform the volume calculation and
LNG carrier with prismatic tanks, depending on the generatereports.
membrane type, the ship should be trimmed to the
best condition to achieve effective stripping and a Priortosuchsystemsbeingenteredintoservice,the
reliable level reading at the end of operations. In calculation, including corrections and gauge tables
general,LNGcarrierswithprismatictanksrequirethe programmed into the system, should be certified as
ship to be trimmed by stern in order to achieve accurate by an independent and competent third
complete or nearly complete unloading of cargo. party,seeSection3.1.3.Oncethesoftwarehasbeen
Dependingonthemembranetype,highertrimlevels verified,thecalculationmethodmay be regarded as
willalso resultin more effectivecargotankstripping reliable so long as the software is unchanged.
providedthatcareistakennottoexceedneithership Software modification may necessitate re
norTerminallimitations,ifanyinplace.Whateverthe certification. It should also be recognized that the
primary membrane configuration, some LNG liquid sensorshavetoberecalibratedatagreedintervalsto
will, however, always remain in the cargo tank and ensurethatinputdataisaccurate,seeSection2.3.5.
that needs to be quantified. It is essential that, as
much as possible during the stripping operation, the

21
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

Custody transfer measurement systems may be


capable of correcting for trim and list both manually The example as illustrated is representative of an
(by operator input of draught readings) and LNG carrier with an older level gauging system. In
automatically (by sensor data). The method should order to calculate the volume transferred, two
be agreed between buyer, seller and terminal conditionsarerequired,knowledgeofthetotalcargo
operator. Trim and list sensors cannot be readily volumeonboard(a)beforethestartofcargotransfer
recalibrated once in situ so manual methods are operations and (b) after these cargo transfer
oftenpreferred.Dueattentionshouldbegiventothe operations are completed. These two conditions are
proper location of the trim/list gauges on board the either manually recorded or directly produced and
vessel. If due to an inappropriate location these printedoutbyacomputerifavailable.Detailsofeach
gauges are influenced by the position of the vessel, cargo tank parameters must be logged. In order to
thismayaffecttheaccuracyoftheinstrument. document a cargo transfer operation, modern
primary custody transfer measurement systems
It is appreciated that a softwarebased calculation produce three printouts, a socalled before
may be less transparent to the buyer, seller and loading/discharge cargo tanks status, an after
terminaloperatorthanmoretraditionalmethods.The loading/discharge cargo tanks status and a third
CTMS as a whole should be dependable in use. In printout just after the after loading/discharge
addition to software verification, due regard should condition; which is usually referred to as the
be given to hardware availability, reliability and Certificate ofLoading/Discharge.Thisthirdprintout
maintainability to minimize the likelihood and summarizes the general parameters of the two
consequenceoffailuresinuse. conditions and shows the volume transferred by
comparing the initial and final cargo volumes. An
It is recommended that instruments are connected example of Custody Transfer Survey documents for
directly to the system, i.e. sensor data cannot be anunloadingofanLNGvesselisgiveninAppendix
manipulated by other systems unless part of the 15.
certified arrangement. Furthermore, computers (PC,
process controllers), data communication links It is essential that the LNG carriers pipework and
(serial, network) and peripherals (screens, manifolds are kept as far as practical in the same
keyboards, printers) should not, in general, be inventory condition at the opening and closing
shared with other applications in order to maintain CustodyTransferSurveys.
data integrity. However, a copy of the calculation
software may be hosted on a sharedworkstation as Figure 6 illustrates the calculation procedure to
abackuptotheprimarysystem. determine the volume of LNG transferred based on
theshipsmeasurements.
Data used in the transfer process, such as tank
levels, may be transmitted from the CTMS to other
ships control and monitoring systems providing this
doesnotinfluencethecustodytransfercalculation.In
general,datashouldnotbetransmittedtotheCTMS
from other systems unless it is part of the certified
arrangement.

Custodytransfermeasurementsystemsaretypically
able to produce printed reports indicating volumes
before and after transfer. However, the ability to
automate this part of the custody transfer survey
doesnotobviatetheneedtoprovide,ifrequired,the
information required to carry out the calculation
manually.

3.3. CALCULATION OF THE


VOLUME OF LNG
TRANSFERRED
Thiscalculationisillustratedbyan examplegivenin
Appendix 5 showing the results of the volume
determination before and after loading the LNG
cargo,withthefollowingassumptions:
a"Gaz Transport"type LNG carrier with5 invar
membraneprismatictanks,
in each tank, one float gauge with a stainless
steelribbon.

22
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

FIGURE 6: FLOWCHART FOR DETERMINING THE VOLUME OF LNG TRANSFERRED

Measured
LevelGauge
level

Temperature
Level Correction Trim
corrections tables List
Density

Corrected
level

LeveltoVolume GaugeTables
conversion

Volume from
tables

Volume Correction Liquid


correction table temperature

Corrected
TANK VOLUME
Volume

tanks

VOLUME
SHIP VOLUME
BEFORE/AFTER
largervolume
smallervolume
(UN)LOADED
VOLUME TRANSFERRED
VOLUME

The tank gauge table also provides the vertical


4. TEMPERATURE heightofthelocationofthetemperaturesensors,so
MEASUREMENT that it may be determined whether the sensor is in
the liquid or in the vapour phase during custody
4.1. LIQUID TEMPERATURE transfermeasurement.

4.1.1. Device The temperature measurement of these probes is
TheLNGtemperatureismeasuredbyprobesplaced converted into degrees Celsius with the help of a
at different heights in the tanks. These probes are data acquisition computer equipped with a printer.
generally three or fourwire platinum resistance Table 1 shows an example of a printout of LNG
temperaturesensors(usuallyPt100),ofwhichthere temperatures when the tanks are filled to 98%
are typically five per cargo tank. Recentlybuilt LNG capacitywithLNG.
ships often have also calibrated backup sensors
fitted next to each main thermal sensor, which can Figure7showsadiagramoftemperaturemeasuring
beselectedandusedincaseofmainsensorfailure. devices installed on a LNG carrier. In this example,
five probes are immersed in LNG in each tank. A
One probe should always be in the liquid phase at sixthprobe(notshown)isinthevapourphaseatthe
thebottomofthetankandonealwaysinthevapour topofthetank.
phaseatthetopofthetank.
The average liquid temperature is calculated upon
the opening and the closing Custody Transfer

23
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

Surveys. The average liquid temperature is


calculated using the temperature reading at each This pressure is needed to calculate the energy of
individual temperature sensor that is in the liquid, displaced gas, see Section 12.1. For this, it is
andnottheaveragetemperatureofeachcargotank. necessaryforthepressuretobeinabsoluteterms.If
the ship's instrumentation measures pressure in
Thermocouples are not used for LNG temperature 'gauge' terms, then the atmospheric pressure must
measurement within custody transfer because they berecordedandaddedtothegaugepressure.
arelesssensitiveatverylowtemperaturesandoften
give a less accurate measurement than platinum Thepressurevalueisrecorded,withtheatmospheric
resistance probes. In addition their installation is pressureifappropriate,atthetimeoftakingtheother
more complex requiring compensation cables. They CTSreadings.
may be installed sometimes for control or simple
indicationpurposessuchascoolingdownorheating Typically the required pressure measurement
ofthetank. accuracy is specified as between 0.1 and 1 % FS
(%offullscaleinstrumentrange).

4.1.2. Testing and accuracy ISO 10976 also addresses vapour pressure
measurement.
The accuracy of the platinum resistance

thermometer varies between 0.1 and 0.2C for

temperaturesrangingbetween145and165C.The
overall uncertainty of the temperature measuring
chain can be estimated at about 0.5C (including
sensor,cable,signalconverter,display).

The sensors are tested for accuracy at the


manufacturers facility prior to installation, and
certified/calibrated upon installation in the cargo
tanks. The temperature measurement is tested and
recalibrated at regular intervals to ensure continued
performance.

The influence of temperature measurement


accuraciesonthedeterminationofLNGdensity(see
Section 15) is important. For instance, for LNG with
3
anaveragedensityintherange440470kg/m ,and
at a temperature around 162 C, the relative
accuracy in density calculation, due to an accuracy
of0.5Cintemperaturemeasurement,isabout0.15
%.

4.2. VAPOUR TEMPERATURE


The temperature in the gaseous phase of the tanks
is used to determine the quantity of gas displaced
during the loading and unloading operations, or the
level correction of the float gauge due to ribbon
shrinkage.

As for liquid temperatures, the average vapour


temperature is calculated using the temperature
reading at each individual temperature sensor, and
not the average temperature of each cargo tank.
However,onlytemperaturesindicatedbyprobesnot
immersedintheLNGareaveragedinthiscase.

Typically an accuracy of1.0 C(intherange 145


to+40C)isrequired.

5. VAPOUR PRESSURE
MEASUREMENT
Vapourpressuremeasurementscanbemadewitha
pressure gauge, which indicates the pressure in the
gasspacesofthecargotanks.

24
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

TABLE 1: EXAMPLE OF TANK TEMPERATURE RECORDINGS (98 % FILLED)

TEMPERATURE TANK 1 TANK 2 TANK 3 TANK 4 TANK 5


PROBE C C C C C

T1 161.83 161.90 161.94 161.89 161.84

T2 161.80 161.88 161.91 161.89 161.86

T3 161.82 161.87 161.92 161.90 161.90

T4 161.79 161.86 161.87 161.88 161.81

T5 161.82 161.82 161.88 161.91 161.84

TANKAVERAGE 161.81 161.87 161.90 161.89 161.85

The liquid temperature in each tank is determined by the arithmetic average of temperatures indicated by the
probesthataredippedintheLNGandthatareinworkingorder.

FIGURE 7: DIAGRAM OF TEMPERATURE MEASURING DEVICES ON BOARD A LNG CARRIER


25
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

gasholder after completion of the sampling


6. SAMPLING OF LNG processforofflineanalysis.

discontinuous sampling: (referred to as


6.1. LNG SAMPLING PRINCIPLES intermittentbyISO8943)Thissamplingprocess
InordertodeterminethequalityoftheLNGitisfirst alsoinvolvescontinuous collection of LNGfrom
necessary to undertake particular operations to the LNG flow line during loading/unloading
conditionthefluidsampledfromitsinitialstate,liquid operations. However, the regasified LNG from
at low temperature, to a final state, gas at ambient thevaporizerisinthisprocesspartlydirectedto
temperature, without partial vaporization or loss of an online GC and partly into a constant
product. pressure floating piston (CP/FP) sample
container (definition according to ISO 8943).A

CP/FP sample container is capable of


Sampling of LNG includes three successive
maintaining constant pressure during the
operations:
sampling of gas from the process line into the
takingarepresentativesampleofLNG, gas cylinder. The gas sample collected in the
CP/FPsamplecontainerisforofflineanalysis.
completeandinstantvaporization,

conditioning the gaseous sample (e.g. ensuring
constant temperature and pressure) before European Standard EN12838 specifies tests to be
transportingittotheanalyzerand/orsampler. carried out in ordertoassessthesuitabilityof these
two LNG sampling systems. However, this EN

standardwaspublishedbeforethecurrentversionof
Sampling is the most critical point of the LNG
ISO8943andactuallyreferstothepreviousversion
measurements chain. Each step must be taken
of this ISO standard. The most important difference
withoutchangingthesamplecomposition.Thisisby
isthatthediscontinuousprocessdescribedintheEN
farthemostcomplicatedphaseofthemeasurements
12838 standard is replaced by the intermittent
andmostoftheproblemsobservedindetermination
process in ISO 8943. The discontinuous method
of the energy loaded or unloaded come from the
directs all sample gas to the online GC while the
sampling system. The sampling system is not
intermittentprocessinadditionincludesamethodto
changeable during unloading/loading. Some
fillsampleCP/FPcontainers.Thisallowsforretained
operatorshaveabackupsamplingsystemtoensure
sampleanalysis.
sample collection in the event of failure of the main
system.
CurrentLNGsamplingsystemsinoperationdeviates
The LNG industry has evolved the sampling from those in the standard and cause confusion
processes. The "spot (discontinuous) sampling whetherthesystemcan bedescribedascontinuous
system"describedinthefirsteditionofthishandbook ordiscontinuous.
has become almost obsolete for CTS
measurements. It is therefore recommended to use The LNG sampling devices shall be such as to
thisonlyasabackupsystemincaseoffailureofthe ensure the total and continuous vaporization of a
main device. The spot sampling system may be quantity of LNG. This quantity must be sufficient for
used, however, for the (manual) sampling for taking gaseous samples representative of the LNG
impurity analyses. For completeness, description of being(un)loaded.
thissystemisgiveninSection6.9.
Most LNG terminal ship unloading lines (and LNG
The sampling processes currently used in the LNG production plant loading lines) have two main
industry are mainly of two types; continuous operatingmodes:
sampling and intermittent sampling, both sampling
processes as defined in ISO 8943. The intermittent circulation of LNG from the storage tanks when
sampling is also referred to as discontinuous thereareno(un)loadingoperations,
samplinginotherpublicationssuchastheEN12838
LNGCarrier(un)loading.
standard and the current GIIGNL handbook. Please
notethattheterminologycontinuous/discontinuousis
differentfromthatusedinthefirstandsecondedition Theoperatingpressurecanbequitedifferentineach
ofthishandbook. operatingmode.

The terms continuous and discontinuous sampling Thesamplingsystemshouldpreferablybedesigned
are related to the analysis of gaseous LNG, that is, in such way to assure representative vaporization
after evaporationofthesampledliquidstream.LNG even during the lowest LNG line pressure. Sub
sampling systems always sample LNG on a cooling (see ISO 8943, annex A) of LNG up to the
continuousbasis: vaporizer should be ensured at all times when the
systemisusedforheatingvaluedeterminationofthe
continuous sampling: LNG. This could require vent gas flow adjustment
This sampling process involves continuous between changes in operating mode or vaporizer
collectionof LNG from the LNGflowline during setting adjustments (temperature, power). Lowering
theloading/unloadingoperations.Theregasified the static pressure should be carried out very
LNG from the vaporizer is thereafter prudentlytoavoidadverse(e.g.geysering)effectsin
continuously fed into the gas sample holder. theselonglines.
Gas sample containers (definition according to
ISO8943)arefilledwiththemixedgasfromthis

26
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

6.2. SAMPLING POINT


Thesamplingpointisgenerallylocated:
on the main loading pipe (line), after the LNG
pumpssendout,
on the main unloading pipe (line), after the
unloadingarms.

LNGmustbesampledduringastableflowofLNG.It
ispreferabletoinstallthesamplingpointascloseas
possible to the custody transfer point (loading arm
flange/ship manifold or main header) so that the
characteristics of LNG do not change before it is
actually transferred because of potential heat input.
However, generally the influence of heat input is
limited,whentheflowdoesnotvarytoomuchinthe
properlyinsulatedmainunloading/loadingline.

Again, please note that the LNG must arrive in a


subcooledconditionatthesamplingpoint/vaporizer.
TheLNGsubcooledconditioncanbedeterminedby
usingthemethodproposedinISO8943(annexA).If
this is not ensured representative sampling is likely
tofail.

27
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

6.3. SAMPLING PROBES

FIGURE 8: EXAMPLES OF SAMPLING PROBES

DirectconnectiononmainLNGline(crosssectionview)

WithaPitottube(transversesectionview)

28
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

In order to keep the LNG sampling flow in a sub 6.5. LNG VAPORIZER AND
cooled condition, the ambient heat input should be
minimised.Thefollowinglayoutscanbeused: CONTROL DEVICES
a straight sampling tube inside the main LNG 6.5.1. Main devices
header(Figure8a), ThevaporizationoftheLNGmustbeascompleteas
adirectconnectionontheperipheryofthemain possible, so that the gas obtained is truly
LNG header by appropriate insulation around representativeofthequalityoftheLNG.
the sampling tube to reduce the ambient heat
inputtothemainpipe(Figure8b). The vaporizer must be designed in order to avoid
fractionation. To that effect heating to a sufficiently
Remark: both above layouts require appropriate high temperature, e.g. 50C, is required to ensure
insulation around the sampling tube, sample stream immediate vaporization of even the heaviest trace
canbefurtherminimisedbyan(optional)permanent components.Apropersamplingconditioningsystem
externalsidestreamLNGflowto a purgelineofthe (accumulator) should be installed to minimize
plant. possible effects from pump pulsation, fractionation
andcondensation.
Improved sampling systems use Pitot tubes. They
canhave: VaporizationofsampledLNGiscommonlyachieved
by direct electrical heating. Alternatively heat
vacuum insulation around the Pitot sampling
exchange with water may be used which is heated
tube (Figure 8c) completed with appropriate
electricallyorbysteam.Thesedevicesaredescribed
insulationofthetoppartoftheprobe,
inmoredetailinSections6.5.1.1and6.5.1.2.
an LNG Pitot sampling tube cooled by a
permanentinternalsidestreamflowinitiatedwith Open tubular atmospheric vaporisers in which the
natural LNG circulation going back to the LNG heat comes from the ambient air have been used.
header(Figure8d). However, theheatflow may not beenoughto avoid
fractionation of the sampled LNG and the use of
these devices is submerged, or steam warming the
coilupdirectly(Shell&tubeLNGSampleVaporizer).
6.4. PIPING ARRANGEMENT This is considered outdated and is not
recommended.
BETWEEN SAMPLING PROBE
AND VAPORIZER
6.5.1.1 Electrical vaporizer
Itisimportanttohavetheliquidsampleline,between
thesamplingprobeandthe LNGvaporizer,asshort In an electrical vaporiser, the sampled LNG passes
aspossible,withasmallinsidediameter(1or2mm through an electricalresistance heated regulator or,
forinstance)andwithsufficientinsulation,sothatthe alternatively, through a tube which is heated
LNGiskeptinasubcooledconditionuntilitreaches electricallybyinduction.
the vaporizer, and this for all possible (un)loading
conditionsincludingrecirculationatlowpressures.

The maximum recommended length of the liquid


sample line between the sampling point and the
vaporizer can be calculated using the following
formula:

W .H
l
q

where:

l = thelengthofpipeline(m)

W = theflowrateofsampleLNG(kg/s)

H = thedegreeofsubcoolingatinletofsample
probe(J/kg)

q = theheatinput(W/m)

In calculating this maximum length of the pipe, the


value of q is the most difficult value to estimate
accurately. The value chosen should be larger than
the design value to take into account the ageing of
theinsulation.

29
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

FIGURE 9: EXAMPLE OF ELECTRIC VAPORIZER



FIGURE 10: EXAMPLES OF WATER OR STEAM VAPORIZERS

a) Withwaterorsteamcirculation

30
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

b) Withwaterwarmedbyelectricalresistance

c) Shell&tubetypevaporizer


Figure 9 shows an example of an electric induction TheLNGsampleflowsinatubingcoilinstalledinthe
vaporizerwhichconsistsofanelectricallyconductive vesselandvaporizesinit.Thecoilisusuallymadeof
tubecoilinwhichLNGflowsandvaporizes,thiscoil stainless steel (recommended) or, sometimes,
being a secondary winding of a transformer. The copper.
primary winding of the transformer is supplied by
electricity and the induced current that develops in One form of steam vaporiser uses a shell & tube
the coil produces the energy necessary to vaporize heat exchanger in which the sampled LNG flows
the LNG sample flow. This heating principle is the through a tube surrounded with a steam or water
only known EN 12838 validated sampling system. heated shell (see Figure 10c). This is considered
The tube coil can be made of copper or stainless outdated.
steel.Stainlesssteelisrecommendedandisusedin
thenewgenerationofelectricalvaporizer.
6.5.2. Auxiliary vaporization control devices
6.5.1.2 Water or steam vaporizer
Control devices must be installed to supervise the
Figure 10a and Figure 10b show diagrams of conditions of vaporization of sampled LNG and to
vaporizers using water circulation (preferably hot protecttheequipment.Theyaretypicallyinstalledas
water),orwaterwarmedbysteamasheatingfluid,or follows.
waterwarmedbyanelectricalresistancesubmerged
inthevessel,orbydirectlowpressuresteam.

31
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

6.5.2.1 On regasified LNG outlet vaporizer outlet temperature not below about
pressureregulator,orgasflowrateregulator,to +20C. (in practice 50C is often used as a set
control the LNG flow to be vaporised point),
independently of pressure or flow rate in the suitable sample flow rates as these depend
mainLNGpipe, highly on the kind of vaporizer used. (inductive
antipulsation bottle, or mixing accumulator, to heated vaporizers use about 1 m(n)/h while
absorbthepressurepulsesandtomaintaingas electrically heated regulators use about
homogeneity, 1500l/h),

impingement chamber, to prevent the sample data to be collected during a steady


entrainmentofpossiblefinedropletsofliquid, (un)loading rate. This means after ramping up,
once full LNG (un)loading rate is reached, and
flow meters and / or flow limiters to control the beforerampingdown.
maximumflowofftake,
During sampling it is essential that the process
pressuremeters,
conditionsofthesamplingsystem(flowtemperature
temperature detection switches, in case of very and pressure) remain in a stable and stationary
high gas temperature (example: no more LNG condition. Sudden changes in gas offtake (for
flow) or very low temperature (example: failure exampleduetotheopeningofavalveortheremoval
ofheatingdevice), of a sample cylinder) that will affect the gas flow,
should be avoided at all times. These changes can
associateddevices,suchasoverpressuresafety cause the gas to fractionate, resulting in improper
valve, electro(magnetic) valve to isolate the samplingandfluctuationinmeasuredheatingvalue.
vaporizer,etc.

6.5.2.2 On LNG inlet 6.5.4. Gas sample collection systems
check valve or restriction orifice, to prevent a
possibleretrodiffusionofvaporisedcomponents To collect either an intermittent or a continuous
tothemainLNGpipe, sample of the vaporized LNG generally two options
areavailable.
needle valve, to control the flow of LNG
3
(however,itisbettertocontrolgasphaseflowin A large gas holder (usually between 0.5 and 1 m )
order not to disturb the state of the LNG maybeusedtostorearepresentative portionofthe
sampled), vaporized LNG during the transfer operation. The
characteristicsofthegascontainedaftercompletion
filter,
andmixingisrepresentativeofthecharacteristicsof
blocking valve to enable safe maintenance of
theLNGloadedorunloaded.Thesegasholderscan
thesampling system byisolationfrom the liquid
be of several types: waterseal type, the sealing
LNG,
waterbeingsaturatedwithgasbybubblingregasified
a liquid bypass system to give full flexibility to LNG through it before filling the holder, (see Figure
thesamplingsystemincaseofmaintenance. 11a) or waterless type with a bladder in the gas
holderandavacuumpump(seeFigure11b),
6.5.2.3 On heating fluid (water, steam or

electricity)
For intermittent sampling a pump can be used that
temperatureregulator,orthermostaticcontrol,to delivers small portions of regasified LNG into a
keepconstantvaporizingconditionsaccordingto CP/FP sample container. The total amount of such
theLNGflow, portions depends on the transfer flow and total
thermometer and thermostats in case of failure amountofLNGcargobeingtransferred.Inthiscase
ofheatingdevices, the sample holder generally has a volume between
500and1000cm(seeFigure11c).
controloftheelectricalsupplyofthetransformer
or of the (submerged) resistance heating A third sampling principle that is also based on
element. intermittent sampling is called spot sampling. With
spot sampling an appropriate quantity of vaporized
6.5.2.4 Safety devices LNGisinjectedinaproperlypurgedsamplecylinder.
Electricalsupplyandelectricalequipmentmustbeof When gas samples are taken in gas sample
a type designed for hazardous conditions (i.e. containers during LNG transfer it should be done
explosion proof). For example, the use of an with regular intervals, depending on the
enclosurepressurizedwithinertgas. characteristics of the transfer lines and equipment,
the organization of operation in the plant, the
durationofgassampleanalysis,etc.
6.5.3. Operating parameters
Example: the common practice for spot sampling is
The following operating parameters can be to take samples during three events only: at 25%,
particularlyrecommended: 50%and75%ofloading/unloadingoperations.

sample pressure at the sampling point greater
than2bars,

32
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

6.6. GAS SAMPLE CONDITIONING 6.6.3. Examples

AfterthesampledLNGhasbeenvaporizedproperly, Examples of general arrangements of sampling


there are two ways of transferring it for analysis: in devicescanbefoundinAppendix6.
gas sample container (definition according to ISO
8943:samplecontainerusedfortheretentionofthe
gas sample and for its transfer to an analyzing
instrument)orbydirectpipingtoagasanalyzer.

6.6.1. Gas sample containers


Thecontainersarefilled:
either at the outlet of a gas sample line coming
fromthegassample holdernormallyreferredto
as a dome sampler, possibly through a booster
compressor or vacuum pump (continuous
sampling, see ISO 8943) (see Figure 11a and
Figure11b),
or directly, at the outlet of the vaporizer unit.
Regasified LNG is fed at specified intervals into
a CP/FP sample container during the transfer
operation using sample pumps, normally
referred to as a piston sampler (intermittent
sampling,seeISO8943)(seeFigure11c),
or directly, at the outlet of the vaporizer unit,
periodicallyduringthetransferoperation.Thisis
referred to as spot sampling, and is not
recommended.(Figure11a,upperpart),

Gas sampling containers should have sufficient


capacity for the analyses that follow, for instance 1
litre.

Theyaregenerallymadeofstainlesssteelandcould
be either with valves at both ends, as shown in
Figure 10d or could be constant pressure floating
piston containers (CP/FP sample container, as
showninFigure11e.

Inview offrequent handling ofthe containers, quick


connectorsarepreferredoverscrewconnectors.

After purging of piping has been performed, several


containers may be filled simultaneously or
successively (not preferred as it is very importantto
have the 2 or 3 bottles filled at the same time),
according to the installed manifold. It is extremely
importanttotakecarethatnoairentersthecontainer
and that the container is sufficiently purged before
takingthegassampletobeanalyzed.

6.6.2. Direct piping to a gas analyzer

A Gas Chromatograph is directly connected to the


vaporiser outlet to perform subsequent analyses
during unloading or loading process providing a set
ofLNGcompositiondata.

In this case, a preferably stainless steel pipe with a


small diameter directly connects the outlet of the
vaporizertoamanifoldattheinletofgasanalyzers.
A set of fittings, regulators, valves, flow meters, etc
are necessary to maintain a constant flow and
pressure. A gas compressor may be required in
order to make up for the pressure drop in the gas
line.

33
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

FIGURE 11: EXAMPLES OF GAS SAMPLE CONTAINERS AND SYSTEMS SEE REFERENCE [14]

a) Watersealtypegassampleholder

b)Waterlesstypegassampleholder

34
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

c)Intermittentsamplingsystem

35
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

d)Metalgassamplecontainer

e)Constantpressurefloatingpistonsamplecontainer

Carryinghandle
Pressure MagneticIndicator
Gauge

Floatingpiston
Valve
Rupturediscsforpressure
protection


6.7. PERFORMANCE OF THE velocity effects on the adsorption/desorption
phenomenon,
DEVICES
lowtemperature LNGvaporizers(atmosphericor
Thefollowingremarkscanbemadeaboutthedesign water at ambient temperature for instance) are
of such sampling devices and the choice of the not recommended because they are susceptible
elementswhichconstitutethem: tofractionation,

to prevent or to limit possible contamination or it is important to ensure that the sampling and
adsorption of the heavy components (C5+) vaporizing parameters remain as constant as
stainless steel is the preferred material for all possibleduringthewholesamplingperiod,mainly
partsincontactwithLNG/NGflowingbetweenthe LNG or gas flow rate, LNG pressure, and
sampling point and the analyzer, since it is less temperatureofvaporization.
reactive than other materials. It is also
recommended that lines, bottles, valves, etc. As the sampling devices have not gone through the
between the vaporizing device and the analyzer same tests, it is difficult to compare their operating
are kept at a constant temperature. During the performances. Examples of calculation of operating
fillingofthegasholderorduringthetransfertothe performances of continuous and discontinuous
analyzer/containers, the gas flow should be sampling devices are given in European Standard
regulated to limit outside temperature and gas EN12838.

36
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted


6.8. SAMPLING PROCEDURE
6.8.4. Sampling parameters
6.8.1. Sampling period
It is important that the operating parameters of the
ItisrecommendedthattheLNGbesampledwhenthe sampling device (pressure, temperatures, flow rates)
LNGtransferflowrateissufficientlystabilized.Soitis are kept as constant as possible throughout the
necessarytoexcludetheinitialperiod,corresponding samplingperiod,inordertoobtainasmoothoperation
tothestartingoftransferpumpsandincreaseofLNG which enables representative and repeatable
flowrate,untilthemainpipeiscompletelyfullofLNG sampling.
anduntilbiphasicoroverheatedLNGcontainedatthe
beginning of the operation has been eliminated, and
until the full flow rate is established. It is also 6.8.5. Utilisation of gas sample containers
necessarytoexcludethefinalperiodwhenLNGflow
ratebeginstodecreasebeforestoppingcompletely. Gassamplescollectedincontainersare:
on the one hand, directly analyzed in order to
When significant changes in pressure or flow rate determine the average composition of LNG
occur in the transfer line, it is necessary to suspend transferred,
sampling temporarily. Sampling can only be
conductedduringstableunloading/loadingflowrate. on the other hand, possibly given to the other
party concerned with the transfer (purchaser or
seller according to the type of LNG purchase
6.8.2. Sampling frequency contract,) or even kept for further investigations,
in case of dispute for instance, during a period
Asfarasfillingofagasholderisconcerned,sampling defined in the contract which may be up to
is continuous during the sampling period, at a fixed severalweeks.
flowrate;spotsamplescanbecollectedinaddition,in
order to control LNG quality and composition and to When the sampling device includes a line whereby
monitor the transfer operation, but the corresponding the regasified LNG is directly piped to the gas
analyses are preferably not used for energy chromatograph, an additional system may be
calculation. designed to collect spot samples (gas sample
container filling station) which are then only used for
When regasified LNG is sent directly to an online control, these samples being taken on a diversion
analyzer,thegassampleanalysisfrequencydepends pipe at the outlet of the vaporizer with the sampling
on the available analyzer (see Section 7.4), but parametersbeingadjustedaccordingly.
typically every 5 to 10 minutes a sample is analyzed
bytheonlineGC.GenerallytheseonlineGCsnever
stop sampling. It is important to mention that for
custodytransferonlythesamplestakenduringstable
(un)loadingoperationsmustbetakenintoaccount.
6.9. SPOT SAMPLING DEVICE
An appropriate quantity of LNG is injected in a
properly purged chamber previously cooled down by
6.8.3. Purging LNG circulation. The chamber is thus partially filled
It is recommended that purging of sample withLNG,andthenisolated.TheLNGsampleisthen
conditioning equipment (line, containers, etc.) is brought to ambient temperature and vaporizes. Thus
carriedout: the regasified LNG fills the whole volume of the
chamber which is designed to withstand the
before starting sampling: corresponding pressure increase. Gas samples are
then withdrawn from the chamber via pressure
introduction of LNG, vaporization and reducingvalvestofillgassamplingcontainers.
circulationofregasifiedLNGinthevaporizer
andpipework, 7. GAS ANALYSIS
subsequent discharge of the purge gas Sampled regasified LNG is analyzed by gas
eithertoatmosphere(smallgasflowrate)or chromatographyinordertodetermineitscomposition.
to the boiloff gas handling system of the
The reason for this is to be able to calculate the
plant,
physical properties needed to calculate the energy

content.Adirectenergycontentmeasurementbye.g.
before filling the gas sample containers:
calorimeterwouldbelesspreciseandwouldalsonot
connectionofthecontainer(s), givethe useful compositional informationtocalculate
otherpropertiessuchasdensityorWobbeindex.
successively filling and emptying each
container (3 times or more) before any gas ThearithmeticaverageoftheonlineGCanalysesor
sampleiscollected, the average composition of the gas chromatographic
analyses of the offline retained samples shall
isolationandremovalofthecontainer(s).
determinethemolarcompositionoftheLNG.Forthe

purpose of determining the molar composition, all


It is advised to keep the sampling system in service
hydrocarbon components heavier than pentane can
between operations, so that the equipment is
be included in the normal pentanes fraction (C5+).
continuously purged and ready for a new sampling
Alternatively, components heavier than normal
withthesameoperatingparameters.
hexanes can be included in the normal hexanes

37
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

fraction (C6+). The notyetnormalized (raw) total carried out against routine calibrations using a
mustbebetween99and101mol%. singlecalibrationstandard.Becausetheassumed
response function can differ from the true one,
Also gas chromatography can be used to determine type 2 analyses can have nonlinearity errors,
someimpuritieslikesulphurcomponentsatlow(ppm) which shall be evaluated by means of a
levels.Adifferentsetupisoftenrequiredthanforthe multipoint performance evaluation carried out in
maincomponents(seeSection7.7). accordancewithISO10723.

Othertraceimpurities,likemercury,requireadifferent The analyzer system must be calibrated or validated
analyticaltechnique.Withmostimpuritiessamplingis priorandaftereachloadingorunloadingoperation.If
critical and special precautions and sampling this calibration/validation fails, a recalibration is
materialsarerequired(seeSection7.7). required.

All classical techniques used to determine the
compositionofgasmixturescanbedirectlyappliedin 7.2.1. Calibration gas/working standard
thecaseofregasifiedLNG.
Calibrationiscarriedoutwith:
Many methods exist in the open literature that help Certified reference gas mixtures (CRMs), i.e: a
technicianstoanalyzeregasifiedLNG,forexamplein reference gas mixture, characterized by a
international standards series, like ISO (e.g. ISO metrological valid procedure for one or more
6974), national institutes like the Energy Institute (IP specifiedproperties,accompaniedbyacertificate
337) or methods from institutes like ASTM (ASTM D that provides the value of the specified property,
1945)orGPA(e.g.GPA2261). its associated uncertainty, and a statement of
metrologicaltraceability.
NOTE: it is important that, the LNG sample to be
analyzed has been vaporized and conditioned Working measurement standard (WMS),i.e: a
correctly to be sure that the sample is truly measurement standard that is used routinely to
representative of the LNG (un)loaded and thus the calibrateorverifymeasuringsystems.
resultoftheanalysisisaswell(seeSection6).
The preparation and certification of the CRM and
WMSmaybeperformedaccordingtostandardssuch
asrespectivelyISO6142andISO6143.

7.1. TYPE OF GAS CHROMATO- The preferred gas mixture used as a calibration gas
GRAPH must include within close percentages all the
components found in the regasified LNG to be
Among the various arrangements that can be found,
analyzed. It is important that all components in the
thefollowingaregivenasexamples:
calibration gas are certified and that this is reported
achromatographwith2or3columnstoseparate on the certificate, e.g. that methane content is not
selectively the components, for instance: one reportedasbalance.
columnforN2,C1toC5andonecolumnforC6+;
or one column for N2, C1, C2 and CO2 and one However,therearealsootherpossibilities:
column for C3, C4 and C5. These arrangements
Gas mixtures could be used that do not contain
were developed in the 1980s and are described
all the components, but which allow the analysis
intheISO,ASTM,GPAandIPmethods,
of the components that are not present in the
alternatively, any modern chromatographic WMS by using one of the elements in it as a
equipment that meets the precision statements reference. This should however not be the case
for all components to be measured in the ISO, fornitrogenandC1toC3.
ASTM,GPAorIPmethods.Atypicalrefinerygas
If different LNG qualities are to be analysed and
analyzerwillfulfiltheserequirements
only 1 WMS is available, this can only be done

after the nonlinearity errors have been

establishedinaccordancewithISO10723.


7.2. CALIBRATION
Theapplicationofgaschromatographyinnaturalgas
isatechniquethatrequirescalibration. 7.3. ENVIRONMENT FOR A GAS
CHROMATOGRAPHIC SYSTEM
Therearetwopossibilities:
Thepracticalrequirementsfortheinstallationofsuch
Type 1 analysis: first determining response a system are the same as those required for any
functions by means of a multipoint calibration highaccuracyanalysisdeviceandmainlyinvolve:
using several calibration standards, followed by
regression analysis. These response functions installation in a closed and temperate (not
are then used to calculate component mole necessarily airconditioned) housing, sheltered
fractions. Type 1 analyses do not involve from the sun, heating sources or draughts.
nonlinearityerrors. appropriate and constant temperature of
calibration gases and sample (injected
Type 2 analysis: assumes a linear response mass/constantvolume),
function, and subsequent sample analysis is

38
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

permanentandsecured electricalsupply,without 7.5. UNCERTAINTY OF GAS


interferences,
ANALYSIS
shroudingand earthconnections ofthe electrical
connectionsbetweenthechromatographandthe The uncertainty of the gas analysis should be
datasystem. calculated according to ISO 69742. Details on the
uncertaintycalculationcanbefoundinSection15.



7.4. ANALYSIS OF REGASIFIED LNG
7.6. RAMAN SPECTROSCOPY
AND RETAINED SAMPLES
At the time of publication of this fourth edition, a
DuringanormalLNGtransferoperation(forinstance, different metrology device for analysing LNG
transfer duration of 12 hours, sampling period composition with direct analysis in the LNG transfer
durationof8hours),thefollowinganalysisprocedures line(s)isundergoingsomepilottesting.
canbecarriedout:
in the case of direct connection between Raman spectroscopy (see also Appendix 7) is an
vaporizer and chromatograph: the analyses can analyticaltechniquethatusesmonochromaticlightto
bemadesuccessivelyduringthewholesampling exciteandidentifythevibrationalmodesofmolecules.
period, with a frequency equal to the duration of Eachmodeofeachmoleculegeneratesashiftinthe
eachanalysisbythechromatographicsystem, frequency of the scattered light. By analyzing the
Example:withoneanalysisevery20minutes,24 frequency and intensity of the scattered light, the
analyses are available during the sampling samples composition may be determined. The
period. For new generation chromatographs, the scattering interaction is so shortlived that the
durationofeachanalysisisreducedto5minutes measurement is independent of the flow rate of the
resultingin96analyses, sample. The technique is viable for all phases of
matter and may be effectively used on mixed phase
in case of periodic filling of sampling containers: samples. Since the intensity of scattered light is
one or more (often two) analyses can be carried dependent on number of molecules that participate,
outsuccessivelyoneachgassamplingcontainer, the best results are achieved with solids, liquids and
with a comparison of results, and possible highpressuregases.
additional analysis or new filling of sampling
containers for important threshold component For LNG, Raman spectroscopy is accomplished
concentrations. This is followed by calculation of through the use of an optical probe that is inserted
the arithmetic average of the percentages of the into a product stream. The probe is typically of
components determined by the analyses stainless steel construction with a small sapphire
considered for the determination of the average window atthetip toallowtheincidentlaser light and
compositionofthesample. scattered light to pass to the analyzer. The probe
Example: two analyses on each container filled must be designed to operate at cryogenic
every hour and calculation of the average, so 8 temperatureswithnolossoffunctiontotheenclosed
averageanalysesduringthesamplingperiod, optics. The light is passed to and from the probe
inthecaseoffillingofagasholder(ISO8943):at through fiber optic cables in a suitably rugged cable
the end of the sampling period, three containers assembly. The most basic Raman analyzer consists
arefilled,oneforeachofthesellerandbuyerand of a laser, spectrograph and a processor to operate
one kept for further investigations (e.g. an them. The laser must be sufficiently stable to allow
independent third party, in case of a dispute). the shift in light to be consistently measured and be
One or more (generally two) analyses can then powerful enough to deliver close to the maximum
becarriedoutonthesamesampleandretained, allowable optical power to the probe tip. The
if there is no significant threshold component spectrographmust also becapable ofmeasuringthe
concentration. frequency and intensity of light to great precision.
Since Raman scattering is a noncontact, non
Inanycase,adatasetisobtainedfromtheregasified destructive technique, calibration may be
LNG analysis. These data have to be subjected to a accomplished without need for custom gas or liquid
statisticalprocesstoeliminateoutliers,badanalyses, samples.Aninstrumentiscalibratedbycharacterizing
etc. With the remaining data an averaging and the wavelength and intensity of the laser and the
normalization treatment is normally done to get the sensitivity of the spectrograph. This can be
best composition (truly representative) of the LNG accomplishedwithstablephysicalreferencessuchas
volume(un)loaded(seeSection8). neongasordiamondcrystals.

From that composition the main LNG properties may The potential of Raman scattering as an analytical
bederived. techniqueforLNGisinits abilitytomeasure a liquid
directly without a phase change to a gas. Raman is
notwellsuitedfortraceanalysisofcomponentssuch
assulphur.

Pilotfieldtestingondifferentlocationsisstillongoing
in view of further validation and eventual acceptance
intheLNGindustry.

39
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

7.7. IMPURITIES CO2 exposed to LNG, turns into a solid similarly


to water (ice) and may block sampling systems
7.7.1. General and damage equipment. Traditionally, as
measured by Davis et al (1962), the solubility of
PossiblecontaminationofLNGisaconcernbecause
carbon dioxide in the liquid and vapour phase
it may have safety and reliability consequences in
has been 340ppm and 85ppm respectively at
relationto:
111.5Kand1bar.Recenttheoreticalresearchby
LNGtransfersystems, Shen & Lin et al (20112012)[15][16] and
experimental research by Gao et al (2012)[17]
thesamplingsystemandanalyticalinstruments, into CO2 solubility in liquid CH4/N2 mixtures at
systemsandequipmentinwhichtheLNGistobe cryogenic temperatures have identified that at
processed, any stage of the liquefaction process it is
estimatedthatonlyslightlymorethan100ppmof
systems and equipment exposed to vaporised carbondioxidecouldbepresent.
LNG.
InestablishingthemajorconstituentsoftheLNG(for
Thisinformationgivenhereisjusttoraiseawareness GHV determination), the danger potentially exists to
of this issue. It is realised that some contaminants ignore(trace)contaminantsand/orimpuritiesbecause
and/or impurities will often be present at some they are normally not present at levels that exceed
(expected) level and the risk of particular tolerance. The consequence may be damage to
consequencesmay dependonthelevel.Itshouldbe equipment,mayleadtosafetyproblemsandpossibly
noted however that some contaminants/impurities resultincustomerrejectionoftheLNG.
maybecumulative.

Corrective and proactive measures may differ from 7.7.2. Specifications and measurement of trace
site to site. The source of contamination may be at impurities in LNG
the location where the LNG is produced, in the
transportcarrierorpossiblyinthereceivingsystemin Trace impurities in regasified LNG which are often
whichtheLNGisprocessedandvaporised. specified in LNG contracts are carbon dioxide,
sulphur components (hydrogen sulphide, carbonyl
Examples of contaminants and their potential impact sulphide and mercaptans) and mercury. The trace
3
include: impurities are normally in the range 01 mg/m (n) or,
3
in case of mercury as low as 5 ng/m (n). Sampling
water: when exposed to LNG, water or water cannot be done in normal cylinders for these trace
vapour turns into a solid (ice) which can block impurities since they are chemically reactive and will
samplingsystems,valvesandinstrumenttapsas beabsorbedbythewallofthesamplecylinder.
wellasdamageequipment,
particulates: metal shavings, welding debris, The determination of trace impurities requires a
insulation, sand, wood and cloth are typical specialapproach.Thiscanhardlybeunderestimated;
examplesofparticulatematerial.Ifinert,themost setup, operation and maintenance are an area for
common problem with particulates would be specialists. All aspects are critical: sampling,
blockagesanddamagetoequipment, calibrationandanalysis.Validationandverificationof
resultsisstronglyadvised beforeusingtheresultsof
sulphur: a sampling point that utilizes copper or theanalysesforcommercialpurposes.
copperalloysmaybedamagedbycontamination
with sulphur and/or the measurement of trace 7.7.2.1 Carbon dioxide
sulphur compounds may be impaired by their The specification limit for carbon dioxide is often
chemicalreactionwithcopper, around 0.01 mol% (100 ppm). The carbon dioxide
mercury: traces of mercury may damage content is normally determined by gas
aluminiumcomponentsbychemicalreactionwith chromatographic (GC) analysis which is capable of
the aluminium. A release of gas due to a analyzingdowntothislimitorevenlower.
resultant failure of the aluminium is an example
ofapossiblesafetyconsequence, 7.7.2.2 Sulphur
Sulphur impurities are normally specified at the
otherhydrocarbons:asamplingorpipingsystem 3
025mg/m (n)level.
that contains, for example, traces of LPG, may

resultinerroneousanalysis,orotherwiseinLNG Sulphur can be specified as total sulphur and/or as
outofspecification.Moreover,duetothelimited
specific sulphur containing components: hydrogen
solubilityofbutanesandhigherparaffinsinLNG,
sulphide (H2S), carbonyl sulphide (COS) and
toohighconcentrationsofthesemayalsosolidify
mercaptans (RSH, where R is an alkyl group; e.g.
and clog sampling systems. Lubricating oil and
methylmercaptan, CH3SH or ethylmercaptan,
sealoilcontaminantsifpresentwillbeintheform
C2H5SH).
ofhardsolids.

Inert gases: nitrogen and air may be present in Sampling for trace sulphur components is not so
bothsamplingsystemsandpipingsystemsfrom, easy; special precautions are needed in order to
perhaps, inadequate or poor purging operations. avoidadsorptionofsulphurcomponentstothewallof
Moreover the presence of oxygen from the air thesampling system devices. Samplingincontainers
maypresentasafetyhazard, ispreferablymadeaccording tothestandardmethod
described in ISO 10715. The interior face of the
sample cylinders must be made out of a material

40
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

which doesn't react with sulphur components. Even commercial equipment is efficient if precautions are
then,thesamplecannotbekeptformorethan8days. taken.

The commercially available materials "silicosteel" or As with sulphur impurities, gaseous sampling for
sulfinertcoatedsteelaresuitableforthisapplication mercury in a cylinder is difficult. It is not reliable for
butareveryexpensive. mercury at low levels which will be adsorbed by the
cylinder wall in a very short period of time causing
7.7.2.2.1 Total sulphur erroneousresults.
Total sulphur can be determined by combustion
techniques, where all sulphur is converted into SO2
which is trapped and quantified. Combustion
techniques are often specified in LNG contracts and
are cost effective but are not the safest to apply. 8. DATA PROCESSING &
Many accidents have occurred due to the rather
violent combustion of gas in an oxygen/hydrogen TREATMENT
flame.
8.1. GENERAL
Newer instrumental techniques like microcoulometry TheLNGsamplingandvaporizationequipmentiskey
(ASTM D 3120), pyrolyses/chemoluminescence, or to Custody Transfer measurements as far as the
hydrogenolysis/ rateometric colorimetry (ASTM D determinationofLNGqualityisconcerned.Theaimof
4045), GC/chemoluminescence (ASTM D5504) or thisequipmentistotakerepresentativeLNGsamples
GC/AES(ASTMD5623)arepreferred. fromtheLNG transferlineandthentoturn the liquid
phase of the sample flow into the gas phase while
7.7.2.2.2 Sulphur components avoidinganyfractionationandselectivevaporization.
In order to determine the sulphur components
separately in a gas, these components must be Then the regasified sample flow is carried through a
separatedfirst.Agaschromatographcanbeusedfor suitable small diameter tubing to a gas
this in combination with a sulphur specific detector. chromatograph for continued gas analysis. Typically
The commonly used detectors (TCD not sensitive anonlineGCisusedforthispurposeasdescribedin
enoughfortracelevelsandFIDnosignalatall)are ISO8943.
not suitable. Several sulphur specific detectors are
available, and are described in ISO 19739 Providedthatthewholesystemworksproperlyandis
(Determination of sulphur compound by GC). correctly calibrated, the gas analysis provided by the
Equipmentforthesemeasurementsisexpensiveand gas chromatograph and its integrator provides the
requires specialist skills. On line measurement is, in molar composition of the LNG and so reflects the
principle,alsopossible. transferred LNG quality. In this way, and based on
thismolarcomposition,relevantLNGpropertiessuch
In principle the total sulphur can also be calculated asgrosscalorificvalueanddensitycanbecalculated
from the components present, under the assumption fromthecomposition.
that all sulphur components are detected and
measured by the GC (which is normally the case, of Modern sampling and vaporization units are
course). complemented with auxiliary devices such as
pressure transmitters, temperature devices, flow
7.7.2.3 Mercury meters,etcforcontrolpurposes.
Mercury can be determined by ISO 6978. For low
levelsofmercurytheprocedureisdescribedinpart2: In general, all signals coming from the auxiliary
Sampling of mercury by amalgamation on devices,gaschromatograph,etc.aretransmittedtoa
gold/platinum alloy. This method is described in the distributed control system (DCS) for control and
range 10100,000 ng/m3(n)(sampling at atmospheric monitoring. That means that a lot of information is
conditions). Mercury is normally in the form of its availableforsubsequentprocessingandtreatment.
metal or light organic mercury components (e.g.
methylmercury).Theseformsaretrappedeffectively During Custody Transfer process a set of LNG
byamalgamation. compositions provided by GCisavailable. Since one
unique LNG composition is necessary for calculating
Mercury sampling is very critical because of the LNGrelevantpropertiesinvolvedinthedetermination
adsorption properties of mercury. Sampling lines of the energy transferred, procedures and guidelines
should be as short as possible (see also ISO 10715 have to be stated and defined to deal with the
for materials, preferably less than a metre in length), availableinformation.
preferably heated (80C) especially if the gas
pressure is high (above some 20 barg) to prevent
possiblecondensationofhydrocarbonsandshouldbe
flushed for a very long time (preferably even 8.2. DATA HANDLING: QUALITY
continuouslyatalowrate,e.g.2l/min).
The molar LNG composition provided by the gas
Commercialequipment is available on the marketfor chromatograph is used to derive LNG properties.
mercury determination but doesn't fit strictly with the Normally density and gross calorific value are
recommendations of ISO 6978 which has been calculated by correlations and procedures from an
developed mainly for high pressure gas averageLNGmolarcomposition(seeSections9and
measurements. However, for mercury levels in LNG, 10). If suitable routines are included in the gas

41
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

chromatograph software it may be used to calculate


these properties directly using temperature as an
9. DENSITY
additionalparameter. 9.1. GENERAL

The set of LNG compositions provided by the gas Therearetwowaysofdeterminingdensity:


chromatograph(s) during the Custody Transfer
process must be managed to produce a the first consists of measuring its average value
representative composition for the whole LNG directly in the LNG carrier's tank by means of
(un)loadingoperation. densitometers,
the second enables the density to be calculated
That is performed in two steps: processing and onthebasisofanaveragecompositionofLNG.
treatment.
Insitu measurement with the help of a densitometer
takes into account the LNG state of equilibrium and
8.2.1. Data processing composition, which means that one no longer
The aim of this step is to eliminate those LNG depends on product sampling and analysis. It would
compositions which were produced by analyses in a therefore seem to be the best method for measuring
period of time during which some operating the LNG density. Unfortunately, technological
parameterswereoutsidepresetlimits. progress has not reached the stage where it is
possible for a reliable apparatus to be available on
Forinstance,when: board a LNG carrier under normal operating
conditions. This is why the second method, which
thesamplingrate is lowerthana preset value of
enables the density to be calculated from the LNG
regasifiedLNGflow(e.g.1.0m3(n)/h),
average composition, is the one that has been
thepressureatthesamplingpointislowerthana selectedhere.
presetvalue,e.g.2bar(gauge),

shipsLNGcargodischargepumpingrateislower
thanapresetvalue,e.g.70%ofnominalrate,
9.2. DENSITY CALCULATION
othercriteriaparticulartotheanalyses. METHODS
Afterperformingthisdataprocessingstep,asubsetof Avarietyofcalculationmethodsexists[6],suchas:
acceptable LNG compositions is ready for data
treatment. stateequationsintheirintegralform,

methodofextendedcorrespondingstates,

8.2.2. Data treatment hardspheremodelmethod,


The aim of the data treatment is to obtain, from a WATSONmethod,
statisticalpointofview,arobustandconsistentresult
that best reflects the quality of the whole transferred ELFAQUITAINEmethod,
LNG. graphicmethodofRCMILLER,
Thisstepconsistsof:
HIZAmethod,
performingastatisticaltestforeachanalysisand
each LNG component in order to determine the revised KLOSEKMcKINLEY method (k1 and k2
presenceofoutliers, tablesinKelvin:K),
evaluating the elimination or not of the detected ISO 6578, also using the revised KLOSEK
outliers(thewholeanalysismustbeeliminated), McKINLEY method (k1, k2 tables in degrees
Celsius:C).
calculating the average composition from the
analysesnotbeingrejected, Validationofthesedensitymodelsbyexperimentation
normalizingthefinalLNGcomposition. isongoing (Ref[18]).Inthis handbook, the preferred
method is the revised KLOSEKMcKINLEY method,
There are different approaches to determine the as described in N.B.S. Technical note 1030
presenceofindividualvaluesinasetofdatathatmay December1980[9]orinISO6578.Itiseasytoapply
be inconsistent and may change the final result: and only requires the LNG temperature and
graphicalconsistencytechniqueandnumericaloutlier compositiontobetakenintoaccount.Thelimitsofthe
tests. These techniquesare explainedinISO 57252 method also encompass the composition of most
standard.Oneofthenumericaltestsrecommendedin LNG produced. Its uncertainty is 0.1%, when either
this standard for dealing with outliers is the Grubbs the nitrogen or butane content does not exceed 5%.
test. Appendix 8 shows the procedure to apply this For these density calculations an electronic
testaswellasanumericalexample. spreadsheetoracomputerprogrammeisoftenused.

Thisdata treatmentstep results in a final LNG molar Comparison between the revised KLOSEK-
composition, representative for the whole LNG McKINLEY method using tables in Kelvin (NBS) and
(un)loading operation, from which the relevant LNG tables in degrees Celsius (ISO 6578:1991) indicates
propertiescanbecalculated. that they are very similar with a relative difference of
-4
about 10 . Tables according to NBS only mention

42
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

components up to C5 however, while tables according


to ISO 6578 mention components up to C6.

10. GROSS CALORIFIC VALUE


10.1. GENERAL
9.3. REVISED KLOSEK-Mc KINLEY
Thegrosscalorificvalue(GCV)ofgascorrespondsto
METHOD (definition according ISO/DIS 6976) the amount of
9.3.1. Limits of the method heat that would be released by the complete
combustionwithoxygenofaspecifiedquantityofgas,
Themethodcanbeusedwithinthefollowinglimitson in such a way that the pressure p at which the
compositionandtemperature: reaction takes place remains constant, and all the
products of combustion are returned to the same
CH4 > 60%mol specifiedtemperaturet asthatofthereactants,allof
iC4+nC4 < 4%mol these productsbeing in the gaseousstate exceptfor
iC5+nC5 < 2%mol water,whichiscondensedtotheliquidstateatt.
N2 < 4%mol
T < 115K Acoherentsetisgenerallyselectedfromthefollowing
referenceconditions:

9.3.2. Formula for the determination of GCV:
This method is based on an empirical evaluation of t = 0Cor15Cor15.56C(60F)or20Cor
themolarvolumeofthemixtureinthethermodynamic 25C,
state of the LNG considered. The density of LNG is
calculatedasfollows: for the volume condition:
p = 1.01325 bar absolute (14.696 psia) or,
Mmix sometimes, 1.01560 bar absolute (14.73
DLNG psia),
Vmix
and

where: t = 0C,15C,15.56C(60F),20Cor25C

Unless mentionedotherwisethereferenceconditions
DLNG = densityofLNG
consideredinthishandbookwillbe:

Mmix = molecularweightofthemixture GCVandWobbe: t =25C
= X i .M i
Mi = molecular weight of pure Gasvolume: p =1.01325barabsoluteand
componenti t =0C,
X i = molarfractionofcomponenti
referred to as "normal conditions", for which the
3
Vmix = molar volume of the mixture expressed in volumeisindicatedinm (n).
l/mol
= X iVi k1 ( k 2 k1 ) .( X N 2 / 0.0425 ). XCH 4 Remark: within the European natural gas transmission
community these reference values have been established as
Xi = molarfractionofcomponenti EASEE-Gas Common Business Practice (CBP) (cfr.
Enclosure 3).
Vi = volume of the component i at the
temperatureoftheLNG Anothercommonsetofconditionsforvolumeiscalled
standard conditions (15C, 1.01325 bara) for which
3
the volume is indicated in m (s). These are
k 1, k 2 = correctionfactors
recommendedbyISO.

TablesforcalculationaccordingtoNBScanbefound Inaddition,theunitofenergywillbekJ,whichcanbe
inAppendix 9.For tablesaccording toISO6578,we then converted to BTU according to the conversion
refertothestandarditself. factortableinEnclosure1.


9.3.3. Example of LNG density calculation
A worked out example of LNG density calculation 10.2. METHOD OF DETERMINATION
according to this revised KlosekMcKinley method OF THE GROSS CALORIFIC
usingmolecularweightofindividualcomponentsfrom
ISO6976isgiveninAppendix10.
VALUE
Thegrosscalorificvaluecanbedetermined:
The calculations can be made e.g. to six decimal
placesandthefinalresultcanberoundedofftothree bymeasurementwithcalorimeters,
decimals. More information on the rounding bycomputationonthebasisofthecompositionof
procedurescanbefoundinSection15. thegasandthereferencedata.

43
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

pv ,Tv = gasvolumemetering
10.2.1. Determination with the help of
conditions
calorimeters

Several types of calorimeters can be used, among R = molargasconstant
them: 8.3144621J/mol/K

manualcalorimeters,
thephysicalconstantsGCVi(mol)andRbeing
deflagrationcalorimeters
specifiedincoherentstandards,e.g.allISO,or
withauxiliaryfluidcirculationexchanger allGPA,.
automaticcalorimeters,
waterflowcalorimeters Note on the exact value of the molar gas constant R.
ISO 6976 : 1995 mentioned R = 8.314 510 J/(mol.K)
aircirculationcalorimeters and referred to CODATA 1986.ISO CD 6976:2013
empiricalcalorimeters, adopted the latest CODATA 2010 value of 8.314 462 1
J/(mol.K).
withindirectmeasurementoftheheat

produced
ii) Real gas GCV (volumetric) =ideal gas GCV
infraredabsorptioncalorimeters
dividedbythecompressibilityfactorZ
sonicvelocitycalorimeters
opticalinterferometertypecalorimeters
X i .GCVi (vol )
Rautergasanalyzers. GCV(vol )
(X i . MVi )Z
A detailed study of these different types of devices
canbefoundinReferences[7]and[8]. where:

These devices enable the volumetric gross calorific Xi = molarfractionofcomponenti
valuetobe measuredbut, asLNGCustodyTransfer
requires the mass gross calorific value of LNG, the
GCVi (vol ) = molar gross calorific value of
densityofLNGmustalsobedetermined.
component i, expressed in
3
Since the composition of LNG is necessary for the kJ/m (n)or(s)
calculationofthedensityofLNG(seeSection9),the
gross calorific value is always calculated from Z = compressibility factor of the gas
composition data as well. This explains why mixture
calorimeters are not used for the purpose of LNG
CustodyTransfer. the values of GCVi (vol) and the method of
calculation of Z being specified in coherent
standards.
10.2.2. Determination of GCV by calculation
ISO 6976 or GPA 2172 standard give methods
10.2.2.1 Examples of formula
for calculation of Z (and respectively ISO 6976
a) GCV (volumetric) expressed in coherent units, orGPA2145forGCV).
3 3
e.g.inkJ/m (n)orinkJ/m (s),canbecalculated,
withoneofthefollowingformulaedependingon
whether the ideal gas or the real gas calorific b) GCV(mass)expressedinkJ/kgcanbe
valuesareconsidered: calculatedbyoneofthefollowingformulae:

i) IdealgasGCV(volumetric) X i . M i .GCVi ( mass )
GCV(mass )
X i . M i
X i .GCVi ( mol )
GCV(vol )
X i . MVi
where:

or Xi = molarfractionofcomponenti
p
GCV(vol ) X i .GCVi (mol ). v
R .Tv GCVi (mass) = massgrosscalorificvalueof
componenti,expressedin
where: kJ/kg

Xi = molarfractionofcomponenti Mi = molecularmassofcomponent
i,expresseding/mol
GCVi (mol ) = molargrosscalorificvalueof
componenti,expressedin thephysicalconstantsGCVi(mass)andMibeing
specifiedincoherentstandards,
kJ/mol


Note: In Japan the above formula is preferred.
MV i = molarvolumeofcomponent

i,expressedinm3(n)/mol or:

44
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

X i .GCVi (mol ) Note: In Japan the above (mass-based) formula is


GCV( mass ) preferred also for gas displaced.
X i . M i

with Xi, GCVi(mol), Mi, MVi as defined in Section
where: 10.2.2.1andYi=molarfractionofcomponentiingas
displaced.
Xi = molarfractionofcomponenti
It should be noted that the molar composition of the
GCVi (mol ) = molar gross calorific value of displaced gas is different from the LNG composition.
component i, expressed in Itisdetermined eitherbygas analysis,or calculated.
kJ/mol Practicalcalculationispossiblewithasimpleformula
based on empirical Ki values for each component. Ki
Mi = molecular mass of component for each component is defined as an empirically
determined ratio between the molar fraction in the
i,expresseding/mol gaseousstateandthemolarfractionintheliquidstate
ofthatparticularcomponent.
the physical constants GCVi(mol) and Mi being
specifiedincoherentstandards.
Yi K i . X i

10.2.2.2 Examples of charts of basic physical For practical displaced gas calculations, these Ki
constants valuesareusuallylimitedtothemostsignificantones
i.e. of the components with the lowest atmospheric
One of the following standards may be used to
provide tables of physical constants and methods of boiling points: nitrogen, methane and sometimes
calculation of factors necessary to determine the ethane. Considering the relatively small energy
grosscalorificvalue: contentofgasdisplacedcomparedwiththeenergyof
the LNG (un)loaded (typically less than 1%), it is

generallyviewedthatthissimplecalculationapproach
ISO6976
ASTM3588 results in a sufficiently accurate overall calculation of
GPA2145 energytransferred.

GPA2172
The calculations in the example show that the
HM21
contribution of ethane vapour to the GCV of gas
These standards, or other publications which are displacedisverymarginal.Thereforeitcanbeargued
that for simplicity it may even be dropped in the
used, refer to the results of works enjoying
international recognition, and except for ISO 6976, calculation of this GCV. Nitrogen is then the only
most of them, at least partly, to the works published component of which the molar fraction should either
by the American Petroleum Institute (A.P.I.), in be measured in the gas returned flow, or be
calculated as shown in the example in Appendix 11,
ResearchProject44.
Table A112. The following typical Ki values for

nitrogen, methane and ethane at near atmospheric


10.2.2.3 Example of calculation pressureandtemperaturesaround160Care:

A worked out example is given in Appendix 11 KN = 23(range:2026)
(TablesA111,A112). 2

KCH = 1
4
The worked out example is based on the use of KC2 H 6 = 0.005
referencetablesandmethodsofcalculationspecified
intheISO6976standard(1995edition).
The Ki values for any other component are assumed
Thecalculationscarriedout,resultinthemassbased tobezero.
grosscalorificvalueusedforthedeterminationofthe
energy of LNG transferred (Table A111, see Inthisway,startingfromthemolarcompositioninthe
Appendix 11), and the volume based gross calorific liquid state determined by LNG sampling and gas
value usedforthedetermination ofthe energy ofthe analysis, the molar composition in the gaseous state
gasdisplaced(TableA112,seeAppendix11). can be calculated. In the event that the sum of the
thusobtainedfractionsdoesnotequal100%,alinear
Theformulaechosenhereareasfollows: correction of the value of each component must be
carriedout.
ForLNG:
X i .GCVi ( mol )
GCV( mass )
X i . M i

Forgasdisplaced
Yi .GCVi (mol )
GCV(vol )
Yi . MVi

45
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted


11. ANALYSIS REPORT
12. ENERGY OF GAS
Astandardanalysisreportcouldbeasfollows:
DISPLACED OR CONSUMED
11.1. IDENTIFICATION TocompletethecalculationofnetenergyoftheLNG
Nameoftheship transferred, the figure derived from the liquid volume
Date calculationmaybeadjustedfor:
(un)loadingterminalandport The gas sent back from the ship during the
Country(originanddestination) loading operation or the gas transferred to the
Beginningofloading/unloading LNG carrier during the unloading operation
Endofloading/unloading expressed in energy terms is determined as
Referencenumber described in Section 12.1. This may be referred
to as 'buyback' gas. This gas is not commercial
natural gas nor vaporized LNG but BOG whose
composition is mainly methane and some
11.2. BASIC DATA nitrogen.
Custodytransferdatabeforeandafter and the gas possibly consumed by the LNG
(un)loading: carrierasfuelintheengineroomorintheships
levelmeasurementsineachcargotank GCU(gas combustionunit)duringthe operation,
temperaturesofLNGanddisplacedgas determined as described in Section 12.2. This
trim,listandthermalcorrectionsineach gascanbeeitherBOGsimilarlytothebuyback
cargotank gasiftheshiptakesitsfuelgasingaseousform
volumeineachcargotank directly from the vapour header, or vaporized
meantemperatureoftheLNG. LNG if the ship takes its fuel gas in liquid form
from its cargo tanks with a small spray pump
meantemperatureofthedisplacedgas
before the LNG gets regasified and injected into
summed volume of all cargo tanks before
the engine/burner (e.g. ships with dual fuel
andafter(un)loading
motors).
Certificateof(un)loading:
totalvolume(un)loaded=difference
betweenvolumesbeforeandafter
(un)loading 12.1. ENERGY OF GAS DISPLACED
cargotankspressure(absoluteand/or FROM THE LNG TANKS
gauge)andcorrespondingmeanpressure
This quantity is determined by the following formula,
Portlog whetheritisreceivedbyorsentbacktotheship.
Noticeofreadiness
273.15 p
Resultsofcontinuous/discontinuoussampling Egas displaced VLNG . . GCVgas
(seeSection6) 273.15 T 1.01325

Numberof(discontinuous)samples. where:

Egas = quantity of energy in gaseous form

displaced during loading or unloading,


11.3. RESULTS expressedinMJ

For the transferred LNG and vapour (displaced gas, VLNG = volume of the LNG loaded or unloaded,
etc.,seeSection12):
expressedinm3

Net(un)loadedLNGvolume(m) p = absolute pressure in the tanks,
Net(un)loadedLNGmass(tonne) expressedinbar
MeancompositionofLNG(mol%),includingnC+
MolecularweightofLNG(kg/kmol) T = mean value of the temperatures of the
Pseudomolarvolume(m/kmol) probes not immersed in LNG, expressed
Correctedmolarvolume(m/kmol) inCelsiusdegrees.
3
(Un)loadedLNGdensity(kg/m )
Gasdensity(kg/m)atspecifiedconditions GCVgas = GCV of the gas in gaseous state
Massgrosscalorificvalue(MJ/kg) contained in the ship's tanks, expressed
Volumegrosscalorificvalue(MJ/m3(n)) 3
in MJ/m (n) or MJ/m(s) in accordance
3
WobbeIndex(MJ/m (n)) withISO6976.
3 3
Expansionratio(m (s)gas/m LNG)
(Un)loadedenergyquantity(GJ) AsindicatedinSection2.3.3,theparametersp,Tand
Vapourreturnenergyquantity(MJ) GCVgas can be either measured, estimated, or taken
Anygastoengineroomenergyquantity(MJ) asconstantsdeterminedbyexperience.Anyresulting
LNGvesselenergyconsumptionduring inaccuracy would only concern the energy of gas
(un)loading(MJ) displaced which represents less than 1% (typically
Net(un)loadedenergyquantity(GJorMMBTU) about 0.3%) of the quantity of energy in the LNG

46
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

transferredandmaythereforebenegligible.InJapan propulsion,thispercentagemayratherbeoftheorder
itiscommonpracticetoassumethereturngastobe of0.04%orevenlower.
100%methaneinthecalculationoftheenergyofgas
displaced. Thispercentagewilldependonseveralfactors:
the ratio of BOG consumption versus total fuel
Forexample: consumptionintheengineroom,
theenergytechnology onboardtheLNGcarrier,
T = 140C refer the Energy Efficiency Design Index
3
GCVgas = 36MJ/m (n) (EEDI),
p = 1.150bar the power and energy management in place on
board the LNG carrier, refer the Energy
and in this case, the formula can be simplified and EfficiencyOperationalIndex(EEOI),
becomes,withenergyexpressedinMJ: thetimeintervalbetweentheopeningandclosing
CTS (e.g. extended if the LNG transfer rate has
Egas displaced = 83.82 . VLNG tobereducedforoperationalreasons),
thesizeoftheLNGcarrier.

The abovementioned typical percentages refer to
12.2. ENERGY OF GAS CONSUMED standard(un)loadingperiodsofsome12to14hours.
Thepercentageofenergyofgasconsumedasfuelby
AS FUEL BY THE LNG CARRIER theLNGcarrierwouldbehigherifthedurationofthe
The LNG carrier, subject to agreement of buyer and LNGtransferoperationisincreased.Butitcouldalso
seller,mayuse gasasfuelin its engine room during belower, forinstance,when a ship is reloaded atan
the loading operation (FOB cargo) or unloading LNG regasification terminal and the ships cargo
operation (CIF or DES cargo), between the opening transferpumpsarenotrunning.Inthiscasetheships
andclosingCTS. GCU(gascombustionunit)canberunningaswellin
order to handle the boiloff gas produced, hence
Thisamountofenergycanbedetermined: reducing the vapour sent back to the shore or ship
beingunloaded.
eitherbythemeasurementofthetotalvolumeof
gas Vg consumed (gas flow meter on board the
LNG carrier) and the evaluation of the GCV of
gasasdescribedinSection12.1forLNGvessels
usingboiloffgasasfuel:
13. ENERGY TRANSFER
DETERMINATION
Efuel gas Vg .GCVgas The activities required for proper custody transfer
duringLNG(un)loadingaredescribedintheprevious
For LNG vessels using forced vaporisation of sections of this handbook. When the measurements
LNGcargo,thequalityoftheLNG(un)loadedcan and calculations have all been completed the net
betaken, energy transfer can be calculated by applying the
formulainSection2.1.
or by a formula agreed by the seller and the
buyer and based on the average gas Table 6 gives an example framework for a quantity
consumption experienced for a given ship at the certificatefortheLNGterminalsandshipsrecordsof
occasionofitsfirst(un)loadingwhilstburninggas anunloadingvariation.
as fuel. This formula may take into account the
duration of the operation, the electrical power
consumptionoftheLNGcargopumpsand/orgas
compressors (blowers) and ballasting pumps in
operation in the time interval between the 14. ENERGY TRANSFER
opening and closing custody transfer surveys MEASUREMENT
(CTS) and the actual average efficiency of
electrical power generators in the engine room. Allelementsenablingthedeterminationoftheenergy
Considering that the amount of energy of gas transferred have now been determined. The formula
consumedintheengineroomissmallcompared for calculating the LNG transferred (mentioned in
with the overall uncertainty of the energy of the section2.1)cannowbeappliedwithintheframework
LNG transferred, parties may agree on a ofTable6givenasanexamplefor:
contractually assumed amount of energy of the QuantitycertificateattheLNGterminal
gas consumed on board during the time interval Unloading / loading certificate for commercial
betweentheopeningandclosingCTSs. use.

For an LNG carrier of 125000 to 145000m LNG
with steam boilers and a steam turbine propulsion
plant, the energy of the gas consumed in the engine
room between the opening and closing CTS may
typically amount to some 0.05 to 0.06% of the total
energyofLNGtransferred.

On the other hand, for an LNG carrier of 155000 to


175000m LNG with dual fuel diesel electric

47
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

Thevolumeuncertaintyofonetankcantypicallyvary
15. UNCERTAINTY OF THE from 0.20 % to 0.55 % (k = 2). To determine the
ENERGY TRANSFER volume uncertainty of the whole LNG carrier, it is
necessary to take into account the correlation
DETERMINATION between the parameters involved: some sources of
The evaluation and expression of uncertainty in uncertainty are highly correlated (tank calibration or
measurement shall be in accordance with the measurement system) whereas others can be
principles established in the ISO document Guide to considered as uncorrelated (each tank has individual
the expression of uncertainty in measurement. The gauges for level, temperature and pressure).
following sections describe the main sources of Nevertheless, it may be difficult to determine the
uncertainty for each variable involved in the energy degree of correlation (positive correlation); thus, a
transfer determination, in order to help users to conservative approach is assumed and the
elaboratetheuncertaintybudgets.Notethatsituations uncertainty of one tank can be used as the
with malfunctions are not considered in the uncertaintyofthewholeLNGcarrier.
calculations.
Note that in 3.01 edition the volumes of each tank
were considered as uncorrelated (independent),
obtaining that the uncertainty of the whole volume
15.1. VOLUME was less than the uncertainty of one tank. In this
There are many measurements involved in the edition, a more conservative approach is used,
determination of the volume of an LNG carrier. The avoidinganunderestimationoftheuncertainty.
uncertainty must take into account all of them,
although some uncertainty contributions are Note for spherical tanks:
negligible. The following table shows the main For spherical tanks, owing to their geometry, the
sourcesofuncertaintyinthedeterminationoftheLNG volume uncertainty, as a result of the small
volume(seetypicalvaluesinAppendix12). inaccuracies of the level gauge, is significantly less
than for prismatic membrane tanks.
Table 2: SOURCES OF UNCERTAINTY IN VOLUME
15.1.1. Cargo Liquid Lines
Source of uncertainty Value The previous section addresses the uncertainty of
Gaugetables CalibrationCertificate volume measurement of the cargo tanks. To have a
Level Note1 full picture, the issue of contents of the cargo liquid
linesneedstobeconsidered.
Liquidtemperature CalibrationCertificate
Vapourtemperature CalibrationCertificate As specified in Section 3.3 and in the ISO standard
List(ifapplicable) CalibrationCertificate ISO 6578, cargo lines should be maintained in the
same condition of pressure, temperature and density
Trim(ifapplicable) CalibrationCertificate (liquid or vapour) during the opening and closing
Tankthermalexpansion
Bibliography
CustodyTransferSurveys(CTS).
factor(Sphericaltanks)
Formost'flatdeck'designsofLNGCarrier(e.g.older
Note 1: The uncertainty of the level measurement membranetype and IHISPB), the arrangement of
shall be determined taking into account the following cargo lines on deck is such that, at completion of
sources of uncertainty: intrinsic uncertainty of the cargo operations, all liquid left in the liquid lines can
gauge,thedeviationofthemean(standarddeviation), drainedbygravitybacktoacargotank.
andtheuncertaintyofthecorrections.
Once the drainage is completed, custody transfer
The volume uncertainty is strongly influenced by the measurementscanproceedasdescribedandthereis
uncertainty of the gauge tables, which can noneedtoconsiderliquidinthecargolines.
significantly vary from one LNG carrier to another.
The uncertainty generally guaranteed by the For all spherical tank designs and for maximized
contractor's calibrations is 0.2 % at ambient membrane type designs where the manifold valves
temperature. A study carried out by the National are below the crossover lines, some consideration is
Bureau of Standards (NBS) [8] on the calibration of neededforundrainableliquid.
the tanks of LNG carriers shows that the real
uncertaintyisfarbetter,andisabout0.05%to0.1 Significant volumes of LNG may remain in the cargo
%. Also according to this study, the systematic manifoldsandcrossoversaftercompletionofdelivery,
uncertaintyduetotheeffectofshrinkageofthetanks typically well over 50 m LNG for the larger ocean
when they have been cooled down should not be going LNG carriers. The fact that this volume often
3
morethan0.07%.Therefore,foratankof26000m , exceedsthetotal accuracyonvolumemeasurement,
3
themaximumguaranteeduncertaintyis52m LNG, underlines the importance of the ships cargo lines
but in practice it is around half of that taking into being in the same condition of density (i.e. liquid or
accounttheNBSstudynotedabove. vapour),duringtheopeningandclosingCTS.Ifthisis
the case, the normal approach is to precool and
In case the LNG carrier is not placed in a well completely fill the cargo lines with LNG prior to the
protected area, the uncertainty due to sagging and first CTS reading on arrival. The assumption is that
hogging should be included in the uncertainty LNGvolumeinliquidlinesisthesame atthetimeof
calculation(onlyformembranetanks). both opening and closing CTS readings, and
thereforecanbeignoredinthecalculation.

48
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

15.2. DENSITY 15.5. GAS DISPLACED


The following table summarizes the sources of Thegasdisplacedrepresentsabout0.3%ofthetotal
uncertaintyinLNGdensitycalculation: quantity transferred and the total uncertainty of the
parameters used in the calculation is 1%. The
Table 3: SOURCES OF UNCERTAINTY IN uncertainty resulting from the calculation of gas
DENSITY displacedintheenergytransferredisthereforeabout
0.003%andcanbeconsiderednegligible.
Source of uncertainty Value

Calculationmethod(KMK) Bibliography

Composition FromISO6974
15.6. GAS CONSUMED IN ENGINE
Valuestakenfrom
Molecularweight
ISO/DIS6976
ROOM
Molecularvolume Bibliography The gas consumedintheLNGcarriers engineroom
Liquidtemperature CalibrationCertificate
duringthetimebetweenopeningandclosingcustody
transfersurveysmaytypicallyamounttosome0.05to
K1,K2 Bibliography 0.06 % of the total energy of LNG transferred (see
12.2)andthetotaluncertaintyoftheparametersused
As most input parameters are correlated to the in the calculation is 1%. The uncertainty resulting
composition analysis, the application of the from the calculation of gas consumed in the energy
conventional uncertainty analysis, stated in GUM, is transferred is therefore about 0.0006% andcan be
complex.Inthesecases,theMonteCarlomethodisa considerednegligible.
practical alternative for the evaluation of the
measurement uncertainty. In case of LNG density, a
Monte Carlo simulation was used for 1000 different
compositions (see Appendix 12), obtaining an 15.7. COMBINED STANDARD
outcomeofthesimulationof0.45%.
UNCERTAINTY AND EXPANDED
UNCERTAINTY OF THE
ENERGY TRANSFER
15.3. GROSS CALORIFIC VALUE DETERMINATION
The uncertainty of the GCV is described in detail in
ISO/DIS 6976 (see equation in Appendix 12), whose The combined relative uncertainty obtained for the
sourcesofuncertaintyare: energytransferredcanbecalculatedas:

Table 4: SOURCES OF UNCERTAINTY IN GCV uE = (uv2+u2+ugcv2)1/2

Source of uncertainty Value The following tablesums upthetypical values ofthe
Composition FromISO6974 expanded uncertainties involved in the determination
Molarmassofeach TablefromISO/DIS of the energy uncertainty (see Appendix 12). These
component 6976 values can be used as reference; however, the
uncertainty budgets should be determined for each
TablefromISO/DIS
GCVofeachcomponent (un)loading.
6976

Source of uncertainty Typical Values (k = 1)


Using the equations as given by ISO/DIS 6976,
typical values for the uncertainty of the mass based Volume 0.10%0.27%
GCVarebetween0.04%and0.07%(k=2). Density 0.23%
GCV 0.02%0.04%

Energy 0.25 0.37 %


15.4. SAMPLING AND VAPORIZATION
Thecombinedexpandeduncertainty(alsosometimes
Theuncertaintyduetosamplingandvaporizationhas calledoveralluncertainty)ofmeasurementwith95%
an impact on the composition measurement and is confidence level (coverage factor k=2) of the energy
therefore influencing both the uncertainties on the transferredis0.50.7%.
GCV and the density. The uncertainty due to
sampling and vaporization is obtained from the Note: if the correlation between density and GCV
repeatability of the analyses, which is already (both variables are obtained from the LNG
included in the calculation of the composition composition) is taken into account (negative
uncertaintyfromISO6974.Thisuncertaintycanbeup correlation in this case: increase of density implies
to 0.3% according to NBS [8] Situations where the decreaseof GCV, and viceversa),the above figures
vaporization fails and the sample is non decreaseslightly.
homogeneousarenottakenintoaccount.

49
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

15.8. ROUNDING OF NUMBERS AND Table 5: DECIMAL PLACES FOR PARAMETERS


COMMERCIAL IMPACT Deci Typical
Parameter Units
According to ISO 800001 (Annex B) rounding mal value
meansreplacingthemagnitudeofagivennumberby Volume(un)loaded m3 0 140000
another number called the rounded number. Two
different rules (rule A and B according to ISO LNGdensity kg/m3 3 450.341
800001)areinuseforthispurposebuttheruleBis
LNGGCV(onamass
ingenerallypreferable(usedincomputers). kWh/kg 4 15.2046
basis)
The greater in magnitude multiple is selected as
theroundednumberifthelastdigitis5. Energy MWh 0 953685
The lower in magnitude multiple is selected as

theroundednumberifthelastdigitis<5.


Examples:
12.24isroundedto12.2 16. SHIP-TO-SHIP LNG
12.25isroundedto12.3
12.26isroundedto12.3
TRANSFER OPERATIONS
Shiptoship (STS) LNG transfer operations enhance
When a number is given without any further the transfer of LNG from one vessel to another
information,itis generallyinterpretedso thatthelast vessel. This can be done directly, in open water, but
digitisrounded.Forexample,thenumber401008is also indirectly by means of intermediate jetty/jetties
assumedtorepresentavaluebetween401007.5and and/orpipinginfrastructure.
401008.4. In this case, the maximum magnitude of
the error in the number 401008 is 0.5. Rounding A basic example of a shiptoship operation is the
shall not be replaced by truncation (i.e, by simply unloading of LNG at a reception place without
cuttingoffthelastdigits)e.g.401007,9becomes401 receiving facility on shore. For this purpose, a FSRU
007. In this case, the number 401 007 represents a (Floating Storage and Regasification Unit) is used.
value from 401 007,0 to 401 007,9 and the error is This vessel receives LNG from another vessel and
smallerthan+1. has regasification facilities on board to regasify the
receivedLNGandtosendthenaturalgastoa(shore)
Rounding of numbers shall always be carried out in gaspipeline.
the last step because rounding in more than one
stage may lead to errors. Besides, accuracy cant For shiptoship LNG transfer operations special
changeduringmathematicaloperations.Ontheother attention is required with regard to the detailed
hand, numerical values of parameters have an analysis of the following aspects: physical vessel
associated standard expanded uncertainty. compatibility, shiptoship communication and cargo
Uncertainty has to be expressed in absolute terms management (in particular sloshing limits, boiloff
with two significant figures (see ISO 800001 for management and cargo transfer rate). Shiptoship
definition of significant digits) ant it sets the limit of operations also require careful monitoring and
significantfiguresofameasure.Thenumericalvalue management of the manifold connection and the
of the property shall be rounded to the least mooring/fenderarrangement.
significant figure in the rounded value of the
expandeduncertainty. ThemethodtoobtaintheenergytransferredinaSTS
operation depends on the contract between the
Forthepurposesof CustodyTransferHandbook,the differentpartiesortheagreed(terminal)rules,butitis
unloaded or loaded energy is calculated with the usualtoapplythestandardrulesfortheunloadingof
followingequation: thedischargingvesselintothereceivingvessel.

E=VLNGxDLNGxGCVLNGEgasdisplacedEgastoER For a shiptoship LNG transfer operation from a
dischargingvesselintoaFSRU,theseller/discharging
Followingtherulesgivenbefore,itisrecommendedto vessel is responsible to provide and maintain the
round off the values of the parameters to obtain an custody transfer measurement systems for volume,
equalnumberofsignificantdigitsforeachparameter, temperatureandpressuredeterminationonboardthe
suchasismentionedinTable5.Inthiscaseboththe LNG vessel and the buyer/receiving vessel is
volume, the mass based GCV, the density and the responsible to provide and maintain the custody
energy have a similar number of significant figures transfermeasurementsystemsattheFSRU,suchas
(6). An example showing the commercial impact of the gas sampling and gas analysis systems.
roundingofnumbersisshowninAppendix13. However, some FSRUs are not fitted with gas
sampling systems to determine the quality of the
received LNG. In this case there are some
alternativestobeconsidered:

50
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

in a position that ensures the probe receives an


- the best option is to calculate the estimated LNG flow representative of the loading flow. If a
LNGqualityreceived.Thisestimationhasto singleloadinglineisused,thenthesampleprobe
takeintoconsiderationtheageingprocessof musteitherbelocatedinmainloadingheader,or
the LNG, the LNG quality originally loaded, theprobemustbeinstalledinthelinewhichisin
the voyage duration and the data of the serviceduringtheoperation. Similarly,the probe
loading Closing Custody Transfer and the should be located at a suitable point in the
STS Opening Custody Transfer. It is process which allows for sufficient pressure to
commonpracticethattheestimationisdone enablecorrectsampling.
byanindependentthirdpartyasforinstance
theindependentsurveyor. Sample probes shall be bidirectional if used in
thereverseflowdirection.
- an alternative is to use a portable testing
apparatus,vaporizerandgaschromatograph Whilst large scale LNG carriers are dedicated to
todeterminethequalitydelivered. LNGservice,smallscalevesselsmaybecertified
for multiproduct service, providing an increased
- the LNG quality originally loaded is risk of contamination to the terminal. Additional
considered as being the LNG quality precautions may be considered to prevent
received. This option does not take into contamination of the BOG, particularly when the
account any ageing of the LNG and hence previous cargo may differ from LNG. It may be
canleadtosignificanterrors. prudent to perform a check of the cargo history,
- thequalityofthegasmeasuredintheshore and where necessary, perform additional
pipeline is considered as the LNG quality sampling prior to loading if there is any doubt
received.Thisisalsonotthebestoption,as aboutthepreviouscargoes.
it is not always possible to determine if the ForvesselsarrivingwithoutanLNGheel,please
gasinthepipelinehasitsorigin100%inthe refertotheguidelinesinsection2.6forgassing
regasified LNG from the STS operation or up and cooling down. In some cases, it may be
whetheritisamixofdifferentcargoes. prudent to perform additional manual sampling
priortoconnectionofthevapourreturnline.
For shiptoship LNG transfer operations from a
(regular)LNGvesselintoanotherregularLNGvessel, The uncertainty figures provided in section 15.1
thecustodytransfermeasurementcanbeestablished forthevolumemeasurementarethosetypicalfor
asfollows: largescaleLNGcarriers.Forsmallscalevessels,
the individual cargo tank uncertainty, the
- STS LNG transfer between two (regular) accuracy of the level measurement, and a
LNGvesselsorbetweenalargeLNGvessel reduced number of cargo tanks, may all
and asmall(er) LNG vessel(splitof acargo contribute to an additional overall uncertainty.
inparcels)onsea,i.e.withoutconnectionto These factors should be considered when
a shore jetty or pipeline. In this case it is determining the overall uncertainty. As the flows
recommended to have a gas sampling and are generally quite small (typically 500 to 1 500
analysissystemonboardofoneoftheLNG m/h) compared to large scale LNG transfers
vessels. Otherwise one of the above (over 10 000 m/h), there might be potential for
mentionedmethodsmightbeused. futureuseofLNGflowmeters(ultrasonic/coriolis)
- STS LNG transfer between two ships to determine the volume transferred, instead of
moored at two jetties of a land based using the cargo tank tables of the small scale
(receiving) terminal. In this case the gas LNGcarrier,seealsoAppendix1.
sampling system installed on the pipeline(s)
of the (receiving) terminal are to be used to
determine the LNG quality transferred. It is
recommended to use the sampling system 18. RELOADING OPERATIONS
whichisinstalledascloseaspossibletothe
unloading vessel in case more than one IN REGASIFICATION
samplingsystemisavailable. TERMINALS

RegasificationterminalsaredesignedtoreceiveLNG
carriers and unload their cargo. Some of them have
17. SMALL SHIP-TO-SHORE recentlybeenadaptedtodiversifytheiractivitiessuch
as for (re)loading operations. However, the
LNG OPERATIONS operationalconditionsduringareloadingoperationat
a regasification terminal may differ from those at a
(Un)loading of small scale carriers brings some liquefactionterminal.
additionalconsiderationswhencomparedtoclassical
large scale shiptoshore transfers. Special attention Particularly, the temperature of the LNG to be
shouldbepaidwhenperformingsmallshipoperations reloaded is usually higher. For safety and efficiency
at a terminal designed for large scale carriers, and reasons, carriers may require to decrease the final
modifiedtoacceptsmallscalevessels. LNG temperature, which is generally achieved by
keeping the pressure low in their cargo tanks by
Care should be taken to ensure proper location sending a significant amount of return gas to the
and type of LNG sampling probes when used in receiving terminal with the ships blower(s). The
reloading operations.The probeshallbe located reloaded LNG will be flashing (excessive boiling) in
as described in section 6.2, but shall be located the ships cargo tanks, resulting in a difference

51
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

betweenthefinalcompositionoftheLNGreloadedin mercaptansulphur
the ship and the LNG quality obtained from the totalsulphur
samplingsystemoftheterminal.Thisphenomenonis oxygen
similar to the natural ageing of LNG during a ships carbondioxide
voyage. The difference between these two LNG nitrogen
compositions may introduce a significant systematic Contaminants
errorintheestimationoftheenergytransferred. particulates
water
This error could be avoided by calculating the LNG mercury
densityandGCVfromtheactualLNGcompositionin oil
the cargo tanks instead of the composition
determined in the terminals pipelines before the
TransferredLNGEnergyQuantityDetermination
ageing. However, as LNG carriers are generally not
Shipprovisions
equipped with LNG composition determination
Shoreprovisions
devices, an alternative to estimate the LNG
Calculation methods (Revised Klosek
composition could be the application of a mass
McKinley,ISO,HIZA,etc.)
balancetotheship,forwhichapropermeasurement

of the gas return flow and quality would be required,
VapourDisplaced,optional:GastoEngineRoom
in addition to the devices already used in the LNG
ConsiderationinQuantityDetermination
CustodyTransferHandbook.
ImpactofdelayofLNGtransfer

Measurement
Shipmeasurementdevices
Shoremeasurementdevices
19. LNG SALES CONTRACT Measurement device accuracy, repeatability
CUSTODY TRANSFER (certificates)
Calibration of measurement devices, analyzer,
CHECKLIST tankgauging
Thefollowingisatabulationofcustodytransferissues Standard,method
toconsiderforinclusioninanagreementforpurchase Frequencyofcalibration
andsaleofLNG. Independentsurveyor
Standards and guides (GIIGNL Custody
DefinitionsandUnitsofMeasure TransferHandbook,ISO,etc.)
Energy units: British Thermal Unit, BTU (or Conditionofship,shoreatgaugingpriortoand
metric equivalent, e.g. kJ, and their multiples, at completion of LNG transfer (e.g. ship's trim
MMBTUorGJ) andlist)
Calorific Value: gross (or net), by volume and Level measurement device specification,
by mass, reference temperature for identifyprimaryandsecondarylevelgauge
combustion, Verificationofconsistencyincomparison
Calorific Value units (BTU/kg, BTU/m (n),
3 Toleranceofdifference
3
BTU/scf,MJ/kg,MJ/m (n),etc.) Pressuremeasurement
Reference temperature and pressure for Measurementdevicespecification
volume(normalorstandardconditions) Eachshiptank
Volume units (normal or standard cubic metre Witnesses to measurement (opening and
orstandardcubicfoot) closingcustodytransfersurveys)
Temperature(degreesCelsius) Methodofrecording
Pressure(Pa,barorpsi) Numberofgaugereadings,timeinterval
Mass(kg) Manual,electronic
Density(kg/m) LNGSampling
Wobbeindex Location
Molefraction Timing,frequency/interval
Significantfigures Flowproportional
Basisforrejection(accuracy)
LNGQualitySpecifications Methodofsamplevaporization
UpperandLowerLimitsfor: Methodofanalysis(gaschromatograph)
GrossCalorificValue(orNet) Locationofanalyzer(i.e.laboratory)
WobbeIndex Reserve samples quantity, retention
Densityattemperature period
Compositionlimits
methane Buyer/Sellerobligationsandrights
ethane Ship tank gauging level, temperature,
propane pressure, list, trim, correction (tables) for
ibutane secondarylevelgauge
nbutane Barometricpressuregauging
ipentane LNGsamplecollection
npentane Boiloffsamplecollection
hexaneplus Sampleanalysis
hydrogensulphide Certification tank gauge tables, instrument
carbonylsulphide calibration

52
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

Witnesscalibration,measurement
Independentsurveyor
Costs
Notification
Recordspreservation

ActionsUponDeviation
Failureofinstruments
Uncertainty in measurements, readings,
calculations
Offspecification

53
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Thirdeditionversion3.01Uncontrolledwhenprinted

TABLE 6: EXAMPLE CERTIFICATE FRAMEWORK

E (VLNG . DLNG .GCVLNG ) Egas displaced Egas to Engine Room


(6)(1)(2)(3)(4)(5)

ENERGY TRANSFER DETERMINATION



QUANTITY UNITS

3
Liquidvolumebeforeloading/unloading m

Liquidvolumeafterloading/unloading m3

3
Liquidvolumetransferred (1) m

3
Density (2) kg/m

Massoftheliquidtransferred kg

MMBTU/kg
Grosscalorific(heating)value (3) GJ/kg
MWh/kg

MMBTU
EnergyofLNGtransferred GJ
MWh
MMBTU
GJ
Energyofgasdisplaced(optional,seeSection12.1) (4) MWh
inkgequivalentLNG
3
inm liquidequivalentLNG
MMBTU
GJ
Energy of gas consumed as fuel on board ship during
MWh
(un)loadingoperation(optional,seeSection12.2) (5)
inkgequivalentLNG
3
inm liquidequivalentLNG
MMBTU
Netenergytransferred (6) GJ
MWh

3
ThequantitiesexpressedinkgequivalentLNGorinm equivalentLNGareusedforcustomspurposesonly.
(Pleaserefertoflowchart,Section2.2.5)

54
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Thirdeditionversion3.01Uncontrolledwhenprinted

ENCLOSURE 1: CONVERSION FACTOR TABLE FOR ENERGY UNITS (1), (2)

MM British Kilowatt- Gigacalorie


Gigajoule 9
Thermal Units hour 10 calories
6 109 joules
10 BTU
(GJ)
(MMBTU) (kWh) (Gcal)

MMBTU 1 1.055056 293.071 0.251996

GJ 0.947817 1 277.778 0.238846

kWh 0.00341214 0.0036 1 0.000859845

Gcal 3.96832 4.1868 1163.00 1

(1) at the same calorific reference temperature; N.B.: all temperatures in accordance with IPTS68 (International
Practical Temperature Scale 1968, cfr. http://www.bipm.org/en/measurementunits/historysi/temperature
scales/),unlessallpartieswouldagreetouseITS90asthereferencetemperaturescale.Pleasenotethatinthe
practical temperature range for LNG custody transfer, the differences between IPTS68 and ITS90 are of the
orderof0.01Korless,andhencenearlyinsignificantandirrelevant,especiallywhenconsideringtheinaccuracy
oftemperaturemeasurementandtheoveralluncertainty(cfr.Section15).

(2) thistableisforinformationonly,pleaserefertoISO1000forfulldetails

55
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

ENCLOSURE 2: LNG AND NATURAL GAS CUSTODY TRANSFER METHODS

ISO standard methods, being the highest level of international methods, are to be recommended. Equivalent
alternatives, as indicated, from other institutes, like GPA, IP or ASTM can also be used, and may be agreed
betweenallpartiesinvolved.ItishoweverrecommendedtoconsistentlyandcoherentlyuseeitherISOorGPA
references or other consistent contractual requirements. In other words, a mix of ISO, GPA and/or other
referencedocumentsshouldbeavoided.

Thesemethodsareregularlyupdatedandreadersarestronglyadvisedtocheckthattheyareinpossessionof
thelatestissues.

METHOD TITLE (alternative method)



General methods
ISO4259 PetroleumProductsDeterminationandapplicationofprecision
datainrelationtomethodsoftest
ISO57251 Accuracy(truenessandprecision)ofmeasurementmethodsand
resultsPart1:Generalprinciplesanddefinitions
ISO57252 Accuracy(truenessandprecision)ofmeasurementmethodsand
results. Part 2: Basic method for the determination of
repeatability and reproducibility of a standard measurement
method.
ISO7504 GasanalysisVocabulary
ISO13686 NaturalgasQualitydesignation
ISO14111 NaturalgasGuidelinestotraceabilityinanalyses
ISO14532 NaturalgasVocabulary
EN1160 InstallationsandequipmentforLiquefiedNaturalGasGeneral
characteristicsofLiquefiedNaturalGas
EN437 TestGasesTestPressuresApplianceCategories
ISO28460 PetroleumandNaturalGasIndustriesInstallationand
equipmentforLiquefiedNaturalGasShiptoshoreinterface
andportoperations.
ISO/IECGuide983 UncertaintyofmeasurementPart3:Guidetotheexpressionof
uncertaintyinmeasurement(GUM:1995)
ISO/IECGuide98 PropagationofdistributionsusingaMonteCarlomethod
3:2008/Suppl1

Sampling LNG/natural gas


ISO8943 RefrigeratedlighthydrocarbonsfluidsSamplingofliquefied
naturalgasContinuousandintermittentmethods
ISO10715 NaturalgasSamplingguidelines
EN12838 InstallationsandequipmentforliquefiednaturalgasSuitability
testingofLNGsamplingsystems

Calibration Gas
Chromatographic
Equipment
ISO6141 GasanalysisRequirementsforcertificatesforcalibrationgases
andgasmixtures
ISO6142 GasanalysisPreparationofcalibrationgasmixtures
Gravimetricmethod
ISO6143 GasanalysisComparisonmethodsfordeterminingand
checkingthecompositionofcalibrationgasmixtures

ISO16664 Gasanalysishandlingofcalibrationgasesandgasmixtures
Guidelines
ISO/TR24094 AnalysesofNaturalGasvalidationmethodsforgaseous
referencematerial

Analysis
Composition/heating
value
ISO69741 NaturalgasDeterminationofcompositionandassociated
uncertaintybygaschromatographyPart1:Generalguidelines
andcalculationofcomposition

56
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

METHOD TITLE (alternative method)


ISO69742 NaturalgasDeterminationofcompositionandassociated
uncertaintybygaschromatographyPart2:Uncertainty
calculations
ISO69743 NaturalgasDeterminationofcompositionwithdefined
uncertaintybygaschromatographyPart3:Determinationof
Hydrogen,Helium,Oxygen,Nitrogen,Carbondioxideand
HydrocarbonsuptoC8usingtwopackedcolumns
ISO69744 NaturalgasDeterminationofcompositionwithdefined
uncertaintybygaschromatographyPart4:Determinationof
Nitrogen,CarbondioxideandC1toC5andC6+Hydrocarbonsfor
alaboratoryandonlinemeasuringsystemusingtwocolumns
ISO69745 NaturalgasDeterminationofcompositionandassociated
uncertaintybygaschromatographyPart5:Isothermalmethod
forNitrogen,CarbondioxideandC1toC5hydrocarbonsandC6+
Hydrocarbons
ISO69746 NaturalgasDeterminationofcompositionwithdefined
uncertaintybygaschromatographyPart6:Determinationof
Hydrogen,Helium,Oxygen,Nitrogen,Carbondioxide,andC1to
C8hydrocarbonsusingthreecapillarycolumns.
ISO6975 NaturalgasExtendedanalysisGaschromatographicmethod
(GPA2286)
ISO10723 NaturalgasPerformanceevaluationforanalyticalsystems

ISO10725 GeneralRequirementsforthecompetenceoftestingand
calibrationlaboratories

Impurities

ISO63261 NaturalgasDeterminationofsulfurcompoundsPart1:
Generalintroduction
ISO63263 NaturalgasDeterminationofsulfurcompoundsPart3:
Determinationofhydrogensulfide,mercaptansulfurandcarbonyl
sulfidesulfurbypotentiometry
ISO63265 NaturalgasDeterminationofSulfurcompoundsPart5:
Lingenercombustionmethod
ISO6327 GasanalysisDeterminationofthewaterdewpointofnatural
gasCooledsurfacecondensationhygrometers



ISO6570 NaturalgasDeterminationofpotentialhydrocarbonliquid
contentGravimetricmethods

ISO6978,part1,2and3 NaturalgasDeterminationofmercury
ISO10101,part1,2and3 NaturalgasDeterminationofwaterbytheKarlFischermethod


ISO11541 NaturalgasDeterminationofwatercontentathighpressure
ISO13734 NaturalgasOrganiccomponentsusedasodorants
Requirementsandtestmethods
ISO19739 NaturalgasDeterminationofsulfurcompoundsusinggas
chromatography

Calculations/conversions
of properties
ISO800001 QuantitiesandunitsPart1:General
ISO6578 Specifiesthecalculationstobemadetoadjustthevolumeofa
liquidfromtheconditionsatmeasurementtotheequivalent
volumeofliquidorvapouratastandardtemperatureand
pressure,ortotheequivalentmassorenergy(calorificcontent).
ISO6976 NaturalgasCalculationofcalorificvalues,density,relative
densityandWobbeindexfromcomposition(referGPA2172and
GPA2145).
ISO122131 NaturalgasCalculationofcompressionfactorPart1:
Introductionandguidelines

57
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

METHOD TITLE (alternative method)


ISO122132 NaturalgasCalculationofcompressionfactorPart2:
Calculationusingmolarcompositionanalysis
ISO122133 NaturalgasCalculationofcompressionfactorPart3:
Calculationusingphysicalproperties
ISO13443 NaturalgasStandardreferenceconditions
ISO15112 Naturalgasenergydetermination
ISO15796 GasAnalyssInvestgationandtreatmentofanalyticalbias
ISO/DIS207652 NaturalgasCalculationofthermodynamicproperties.Part2:
Singlephaseproperties(gas,liquid,anddensefluid)for
extendedrangesofapplication.

Quantity related methods
ISO8310 Refrigeratedhydrocarbonandnonpetroleumbasedliquefied
gaseousfuelsGeneralrequirementsforautomatictank
thermometersonboardmarinecarriersandfloatingstorage
ISO8311 RefrigeratedlighthydrocarbonfluidsCalibrationofmembrane
tanksandindependentprismatictanksinshipsPhysical
measurement


ISO10976 RefrigeratedlighthydrocarbonfluidsMeasurementofcargoon
boardLNGcarriers

ISO181321 Refrigeratedhydrocarbonandnonpetroleumbasedliquefied
gaseousfuelsGeneralrequirementsforautomatictankgauges
Part1:Automatictankgaugesonboardmarinecarriersand
floatingstorage
ISO181322 RefrigeratedlighthydrocarbonfluidsGeneralrequirementsfor
automaticlevelgaugesPart2:Gaugesinrefrigeratedtype
shoretanks
ISO28460 PetroleumandnaturalgasindustriesInstallationand
equipmentforliquefiednaturalgasShiptoshoreinterfaceand
portoperations

58
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

ENCLOSURE 3: OTHER RELEVANT STANDARDS & REFERENCES

1. GPA 22612013, Gas chromatography: "Analysis for Natural Gas and Similar Gaseous Mixtures by Gas
Chromatography"

2. GPA21722009,CalculationofGHV,RelativeDensity,Compressibilityandtheoreticalhydrocarbonliquid
contentforcustodytransfer

3. GPA21452009,Tableofphysicalpropertiesforhydrocarbonsandothercompoundsofinteresttothe
naturalgasindustry(tobeusedwithGPA2172)

4. ASTM D 6667, Standard test method for determination of total volatile sulphur in gaseous hydrocarbons
andliquefiedpetroleumgasesbyultravioletfluorescence

5. ASTMD1945,Standardtestmethodforanalysisofnaturalgasbygaschromatography

6. HM 21 Calculation procedures for static and dynamic measurement of light hydrocarbon liquids (LNG,
LPG,ethylene,propyleneandbutadienes(replacesIP251/76)

7. ISO800001,Quantitiesandunits,Part1:General;AnnexBRoundingofnumbers

8. ASTM E 29 13, Standard Practice for Using Significant Digits in Test Data to Determine Conformance
withSpecifications

9. GeneralENstandardsforLNGfacilities:
CENEN1473InstallationandequipmentofLNGdesignofonshoreinstallations
EN1474part1,2and3,installationandequipmentforLNGDesignandtestingofmarine
transfersystems
CEN EN ISO 28460 Petroleum and natural gas industries Installation and equipment for liquefied natural
gasShiptoshoreinterfaceandportoperations

10. INCOTERMS2010,definingdutiesofsellerandbuyeramongothers:
FOBFreeonBoard(namedportofloading)
CIFCost,InsuranceandFreight(namedportofdischarge)
DESDeliveredExShip(portofdischarge)(INCOTERMS2000)

11. JCGM100:2008EvaluationofmeasurementdataGuidetotheexpressionofUncertaintyin
Measurement.

JCGN200:2012InternationalVocabularyofmetrology.Basicandgeneralconceptsandassociatedterms(VIM)

WEBSITES FOR FURTHER INFORMATION:

http://www.aga.org:AmericanGasAssociation
http://www.eurogas.orgEuropeanNaturalGasCompanies
http://www.gie.eu.com:EuropeanLNGTerminaloperators(>selectgle)
http://www.gas.or.jp/http://www.gas.or.jp/english/index.html:JapanGasAssociation(JGA)
http://www.giignl.orginformationonGIIGNL
http://www.igu.orginformationonIGU,theInternationalGasUnion
http://www.iso.chinformationonISOstandards
http://www.sigtto.org information on SIGTTO LNG shipping and shipshore best safety practices, with
numerousfreedownloadsonsafepractices,esp.forLNGshipshoreinterface.
https://easeegas.eu/#:EASEEGASwebsite,withinformationoncommonbusinesspractices(CBP)inthe
Europeannaturalgasindustry,e.g.referencevolume,referencetemperaturesforgasvolumeandcalorific
value,harmonizedunitsforcalorificvalueandWobbenumber.
http://www.naesb.org/: website of the North American Energy Standards Board, with information on
commonstandardsintheNorthAmericannaturalgasindustry.

59
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

ENCLOSURE 4: NOTES ON TERMINOLOGY

Thetermheatingvalueissynonymouswithcalorificvalue.

The qualifiers for calorific value higher, upper, total and superior are, for the purposes of this
handbook,synonymouswithgross.Likewiselowerandinferioraresynonymouswithnet.

The grosscalorificvalueisused generallyforLNGtradingandwidelyinthe gasindustry.Someusersof
naturalgas,notablypowergenerators,preferthenetcalorificvaluewhichislowerbyroughly10%dueto
itsexclusionoftheheatofcondensationofthewaterproducedbygascombustionwithair.

Specificgravityissynonymouswithrelativedensity.

WobbenumberissynonymouswithWobbeindex.

Compressibilityfactorissynonymouswithcompressionfactor.

Unloadingoperationissynonymouswithdischargingoperation

60
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

LIST OF FIGURES

page
Figures1&2: CorrectionsforthepositionoftheLNGcarrier 14
Figure3: Electricalcapacitancetypelevelgauge 18
Figure4: Floattypelevelgauge 19
Figure5: Radar(microwave)typelevelgauge 20
Figure6: FlowchartfordeterminingthevolumeofLNGtransferred 23
Figure7: DiagramoftemperaturemeasuringdevicesonboardaLNGcarrier 25
Figure8 Examplesofsamplingprobes 28
Figure9: Exampleofelectricvaporizer 30
Figure10: Examplesofwaterorsteamvaporizers 30
Figure11: Examplesofgassamplecontainersandsystems 34
FigureA2.1: Lasertypecargotanklevelgauge 65
FigureA7.1: GraphicalrepresentationoftheRamaneffect 81
FigureA7.2: GCandRamanBTUdatacomparisonforshipunloading 81

61
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

LIST OF TABLES

page
Table1: Exampleoftanktemperaturerecordings(98%filled) 25
Table2: Sourcesofuncertaintyinvolume 48
Table3: Sourcesofuncertaintyindensity 49
Table4: SourcesofuncertaintyinGCV 49
Table5: Decimalplacesforparameters 50
Table6: Examplecertificateframework 54
TableA41 Exampleofgaugetablewithzerolistandtrim 68
TableA42 Exampleofbottomfinegaugetable 69
TableA43 Exampleoftopfinegaugetable 70
TableA44 Exampleofvolumecorrectionsaccordingtotankservicetemperaturesfor 71
selfsupportingtanks
TableA51 VolumedeterminationbeforeloadingLNG 72
TableA52 VolumedeterminationafterloadingLNG 73
TableA53: Exampleofgaugecorrectionsforlowtemperatures 74
TableA54: FloatgaugesExampleoffloatbuoyancy/DIPcorrections 75
TableA55: Exampleoflistcorrectiontable 75
TableA56: Exampleoftrimcorrectiontable 76
TableA81 CriticalvaluesforGrubbstest(ISO57252) 83
TableA82 ExampleGCanalyses 84
TableA83 ExampleGpstatistic 84
TableA84 ExampleG1statistic 84
TableA85 ExampleAveragecomposition 85
TableA91 Componentmolarvolumes 86
TableA92 Volumecorrectionfactork1.103 86
TableA93: Volumecorrectionfactork2.103 87
TableA101 Calculationofmolecularweighting/mol 88
TableA102 Calculationofthemolarvolumeinl/molbytemperatureinterpolation 88
TableA103 Calculationofthecorrectionfactork1inl/moltemperatureinterpolation 89
TableA104 Calculationofthecorrectionfactork2inl/moltemperatureinterpolation 89
TableA111 Grosscalorificvalue(kJ/kg) 91
3
TableA112 Grosscalorificvalue(kJ/m (s))forEgasdisplacedcalculation 92
TableA121 Compositionanalysis 93
TableA122 Sourcesofuncertaintyinvolume 94
TableA131 Commercialimpact 96

62
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.0Uncontrolledwhenprinted

REFERENCES

[1] Le volume du gaz naturelliqufi AssociationTechnique del'IndustrieduGazen France F.


DewerdtetB.CorgierJanvier1987.

[2] GaugetableoftheLNGcarrierMethaniaissuedbyMetrologicalServiceBrussels.

[3] GaugetableoftheLNGcarrierMouradDidoucheissuedbytheMetrologicalServiceParis.

[4] CalibrationofcontainersandgaugesJournaloftheInstituteofPetroleumVolume58n561
1972.

[5] Gas quality Proceedings of the congress of "Gas Quality Specification and measurement of
physicalandchemicalpropertiesofnaturalgas"GroningenTheNetherlands2225April1986
EditedbyGJvanKossum.

[6] La masse volumique du gaz natural liqufi Association Technique de l'Industrie du Gaz en
FranceF.DewerdtMars1980.

[7] Le pouvoir calorifique du gaz naturel liqufi, par M. F. Dewerdt Association Technique de
l'IndustrieduGazenFranceMars1983.

[8] LNGmeasurementNBSIR853028Firstedition1985.

[9] Four mathematical models for the prediction of LNG densities NBS Technical Note 1030
December1980.

[10] Appendix2Informationsource:InvensysFoxboro,Canada.

[11] Appendix 7Informationsources:Scientific Instruments,Inc.,USA, KaiserOptical Systems,Inc.,


USA,andDistrigasofMassachusettsLNGTerminal,Everett(Boston),MA,USA

[12] Appendix 14 Information sources: Case 1: Fluxys LNG Terminal, Zeebrugge, Belgium; Case 2:
th
presentationatSIGTTO59 GPCMeeting,20090422,Boston,MA,USA.

[13] Appendix15Informationsource:ISO10976workingdraft#8

[14] Figures11cand11e:adaptedfromISO8943

[15] CalculationofcarbondioxidesolubilityinliquefiednaturalgasT.Shen&W.LinInternational
JournalofChemicalEngineeringandApplicationsVol.2,no.5,pp.3663712011

[16] Determination of CO2 solubility in saturated liquid CH4+N2 and CH4+C2H6 mixtures above
atmosphericpressureT.Shen,T.Gao,W.Lin&A.GuJournalofChemicalEngineeringData
vol.57,pp.229628032012

[17] ExperimentaldeterminationofCO2solubilityinliquidCH4/N2mixturesatcryogenictemperatures
T.Gao,T.Shen,W.Lin,A.Gu&Y.JuIndustrial&EngineeringChemicalResearchvol.51,
pp.940329752012

[18] M.Richter,R.Kleinrahm,&R.Span.NewApparatusforAccurateMeasurementoftheDensityof
Liquefied Natural Gas (LNG). International Gas Union (IGU), IGRC 2011 Proceedings and
PresentationsPosterPapersSession4.

[19] ISO 57252 Accuracy (trueness and precision) of measurement methods and results. Part 2:
Basicmethodforthedeterminationofrepeatabilityandreproducibilityofastandardmeasurement
method.

[20] Extension of sample sizes and percentage points for significance tests of outlying observations,
FrankE.GrubbsandGlennBeck.TechnometricsVol.14,No.4,November1972.

63
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

APPENDIX 1: USE OF IN-LINE MEASUREMENT OF LNG FLOW

IN-LINE FLOW MEASUREMENT OF LNG

SeveraltypesofLNGflowmeterexistbutLNGshipcargotransferisconventionallymeasuredbyvolumetricgauging
forcustodytransfer.However,thequestionisoftenasked,caninline LNG flowmeasurementbeusedforcustody
transfer,orasacheckmeasurementorforallocationpurposes?Thissectionisintendedtoaddressthattopic.

Performance

It has been demonstrated that both ultrasonic and Coriolis meters can be used for metering LNG at cryogenic
temperatureswithcustodytransferqualityrepeatability.LNGhasanapproximaterelativedensityof0.5andviscosity
of 0.1 cP. From a flow measurement perspective, this application presents few challenges provided the meter can
physically handle the cryogenic temperatures and no flashing (2 phase flow) occurs. Testing at a working LNG
loading facility has shown that a Coriolis mass meter will agree with an ultrasonic meter on a mass/inferred mass
basis within the published uncertainties of both devices. Results from these tests demonstrated repeatability of
0.125%bymass,astepchangeimprovementinrepeatabilityfromshipmeasurement.

IndustrystandardsalreadyexistforliquidmeasurementforbothCoriolisandultrasonicliquidmeters,althoughthey
donotspecificallyaddresscryogenicapplications.Technologyselectioncanbeevaluatedbasedonthepressureloss
available and flow rates required. Coriolis meters which measure mass flow directly dont require a density
calculation,buthavesizeandpressuredroplimitations.Ultrasonicmetersrequireadensitycalculation(e.g.Klosek
McKinleyorCostald),buthavenosizeorpressuredroplimitations.

Given the demonstrated performance in LNG service, either technology could be used for allocation or check
measurement.Itshouldbenotedthatonemajorjointventureisusinginlinemeasurementforallocation(notcustody
transfer)ofLNGproducedandcommingledinacommonstoragefacility,andtwootherLNGfacilitieshaveinstalled
inlinemetersascheckmetersinordertoverifytankerloadingmeasurements.

Proving

With respect to the use of meters for LNG custody transfer, the hurdle is finding an industry accepted practice for
provingameterforuseinLNGservice.AnacceptedpracticeforLNGprovingtoatraceablestandarddoesnotexist
atthetimeofthispublication.Whileinlinemetersoffercustodytransferqualityrepeatabilityintherangeof0.125%
by mass, a meter factor is still required to achieve traceable accuracy. Without a prover, a meter could have a
significantbiases.LNGprovinghasseveralhurdlesthatmustbeovercome:compensationfortemperaturechanges,
capacity,eliminationofboiloff,anddensitychanges.

Loading and offloading of ships can involve volumetric rates of 5000 to 15000 m3/hr, so much like gas meters,
onsite proving doesnt seem practical, and a feasible concept has yet to be developed. A 0.50C change in the
measured temperature can cause a 0.17% change in the calculated density. These challenges would impact the
design of an LNG prover regardless of the physical principles to be employed, e.g. whether it would be based on
mass(weightoftank),orvolume(tankorpiston).Custodytransferusinginlinemeasurementwillnotbecomereality
until an accepted practice for proving is developed and established in commercial use. Once it has, inline
measurementwillbeanacceptablealternativeorabackuptoshipmeasurement.

64
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

APPENDIX 2: LASER TYPE CARGO TANK LEVEL GAUGE

New technology, for information only - see Reference [10].




FigureA2.1:Lasertypelevelgauge

Principle

The principleof lasertypelevelgaugeisbased ona dualtransmitter(opticalhead) composed of an emitter and a


receiver, mounted on the top of the cargo tank. The emitter sends a laser beam down towards the surface of the
liquid; the signal isreflected from the surface, detected by the receiver cell and send back to the unit control. The
systemiscalibratedatseveraltargetlevelsofthetank.

Thefirstcommercialapplicationandcertificationofthisnewlaserbasedtechnologywasinlate2005followingaR&D
effortonboardthevesselLeTellier.NKKKcertificationforthesystemwithanaccuracyupto7.5mmwasgranted
inNovember2005.

Accuracy

Measurementrange : 050m
Resolution : 1mm
Accuracy : 7.5mmorbetter

Features

Onlineaccuracyverification
Temperature,pressure,trimandlistcorrected
Serialcommunication
Nointankequipment

65
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

APPENDIX 3: SOME RECOMMENDATIONS FOR PARTIAL RELOADING OF CARGO TANKS OF LNG


CARRIERS, WITH REGARD TO BOTH SAFETY AND CUSTODY TRANSFER ISSUES.

For information only

Introduction and context

InthelastdecadewehaveseenaworldwidedevelopmentofshorttermandspotcargoLNGtrade,involvingtwo
newoperatingtrendsinLNGshipping:

LNGShippersareusingLNGcarriersforspotcargoesand/orasfloatingLNGstorage,
severalLNGShippershaveconsidered,andsomehavecarriedout,partialunloadingand/orpartialreloading,
i.e.unloadingorreloadingonlyoneorsomeofthecargotanksoftheLNGcarrier.

Whenperformingsuchoperations,dueattentionshouldbegivento:

safeship/shoreoperatingpracticesandprocedures,and
properLNGship/shorecustodytransferprocedures.

Oneoftheconsiderationsforsafeoperatingpracticesandproceduresshouldbetopreventthecreationofstratified
layers and the risk of a subsequent rollover in the cargo tank(s) involved. Stratification and rollover could occur
eitherwhentheLNGcarrierisalongsidetheberthofanLNGterminaloralongsideanotherLNGcarrierundergoing
apartialreloadingoperation,whenatsea,orwhenunloadingitspartiallyloadedLNGcargoinanotherLNGcarrier
oratanotherLNGterminal.Inthelastcase,theriskofstratifiedlayersandsubsequentrolloverisinthereceiving
LNGterminalsshoretanks.

With regard to safe operating practices, both GIIGNL and SIGTTO published some rules of good practice in this
respect:
GIIGNL:RolloverinLNGStorageTanks(19811983),Section3.5Stratificationanddensitydifferencesin
LNGShipsandConsequencesforShorebasedTanks(p.2627).
GIIGNL:GIIGNLLNGRolloverStudy(update2014)
SIGTTO:ROLLOVERPREVENTION,AReviewofCauses,MethodsofPreventionandDamageLimitation
Measures(November1993),seetheparagraphonShipCargoesonpage11.
SIGTTO:GuidanceforthePreventionofRolloverinLNGShips,firstedition2012.

Custody Transfer recommendations when considering the reloading of LNG carriers for partial reloading
operations

StratifiedlayersincargotanksofLNGcarrierscan(andhave)subsequentlyresultedinrollovereitherinthecargo
tanksoftheLNGcarrierorinthereceivingonshoreLNGtanks(seeAppendix14).

Itshouldbenotedthatatthetimeofwritingthereisstilllimitedknowledge,reportingandfieldexperienceregarding
stratification and rollover issues in cargo tanks of LNG carriers. Persistent suggestions that ship movements and
sloshingoftheLNGcargowouldalwaysensurecompletemixingofanystratifiedlayershavebeendisprovedbya
fewreportedcasesofstratificationandevenrolloverincargotanksofLNGcarriers.Thereforeallpartiesactingas
prudent and reasonable operators both offshore and onshore, should adhere to the following guidelines,
particularlywithregardtopartialloadingofcargotanks.

Inviewoftheabove,itshouldbestronglyrecommendedtofirstunloadapartiallyfilledcargotank(i.e.acargotank
filledwithmorethanaminimalheelofonlyafew%ofcargotankheight),andthentoreloaditwithafreshcargoof
LNG.

IfanLNGcarrier,typicallyinaspotcargosituation,arrivesatanLNGterminalwithoneorseveralcargotanksfilled
andothersempty(perhapsexceptforaminimalheel),withtheintentiontoloadtheremainingcargotanksaswell,
thenpriortoagreeingtotheLNGcarrierscallatthatLNGterminalforsuchapartialreloadingoperation,theLNG
ShippershallprovidetheLNGTerminalOperatorwithanappropriatecargotankmanagementprocedureforsuch
partialreloadingoftheLNGcarrier,therebyensuringthatthefollowingrulesforpropercustodytransferarestrictly
adheredto.

Hereareafewproposeditemsforsuchaprocedure.

A. FillinganLNGcargotankwithaheelheightexceeding5%ofmax.cargotankfillingheightshallnotbe
allowed.

66
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

B. Likewise, topping up a cargo with LNG transferred from an LNG Terminal tank shall not be allowed.
Therefore, before starting a partial loading operation for one or several of the ships cargo tanks, if
necessary each cargo tank in either of the abovementioned conditions shall first be emptied to the
recommendedminimumheel level,typicallyonly 0.5or1m ofliquidlevel.Partiesshouldagreeonthe
limitsinLNGtemperaturedifferenceandLNGdensitydifferencebetweentheheelandtheLNGcargoto
beloadedasdefaultvaluesamaximumacceptabledifferenceof1Cand1kg/mrespectivelybetween
thisminimalheelandtheloadedcargoareproposed.

C. If an LNG Shipper proposes to call at an LNG terminal with an LNG carrier with one or several cargo
tanks full and the others empty (or with a minimum heel), and wishes to reload only the empty cargo
tanks,thenthefollowingstrictguidelinesarerecommended,toensureapropercustodytransferbothon
departureandarrival.

FirstofalltheShippershoulddemonstrateinfulldetailandinwritingthatheisabletodeterminewithan
overalluncertaintybetterthan1%(seeSection15)therelevantphysicalpropertiesoftheLNGcargothat
isonboardtheLNGcarrier.

AssumingthattheLNGShipperisabletodemonstratetheabove,thenthepartiesshouldverify,basedon
wellestablishedandprovenmethods,thatthereissufficientcompatibilitybetweenthephysicalproperties
of,ontheonehand,theLNGcargoonboardincertaincargotanks,andontheotherhandtheLNGbatch
fromLNGshoretanksthatistobeloadedintheothercargotanks.Asaguidancethefollowingcriteriaare
recommended:

Max.1kg/mdifferenceinLNGdensitybetweentheLNGparcelsintheshipscargotanks
Max.0.5Cdifferenceinaverageliquidtemperaturebetweentheshipscargotanks
Max.50mbardifferenceinsaturatedvapourpressurebetweentheLNGparcelsintheshipscargotanks.
Thesaturatedvapourpressureineachcargotankshallbecalculatedwiththesamemethod.

Ifthesecriteriacannotbemet,thentheLNGShippershouldrefrainfromsuchapartialreloadingoperation
andfirstunloadallLNGonboard,andnextloadafreshandhomogeneouscargoofLNGfromtheshore
terminal.

Likewise,iftheLNGbatchthatistobereloadedfrommorethanoneLNGshoretank,asaguidancethe
followingcriteriafortheseLNGshoretanksarerecommended:

Max.1.0kg/mdifferenceinLNGdensitybetweentheLNGshoretanks
Max.0.5CdifferenceinaverageliquidtemperaturebetweentheLNGshoretanks
Max.50mbardifferenceinsaturatedvapourpressurebetweentheLNGshoretanks;Thesaturatedvapour
pressureineachshoretankshallbecalculatedwiththesamemethod.

One ofthepurposes ofthese recommendationsisto ensurean accurate custody transferonunloading,
sinceevensmalldifferencesinliquidtemperatureofonlyafewtenthsofaCbetweencargotankshavea
detrimental influence on the combined standard uncertainty and overall uncertainty of the calculated
energytransferred(seeSection4.1.2andSection15).

67
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

APPENDIX 4: EXAMPLE OF GAUGE TABLES (see Section 3.1)

TABLE A4-1: EXAMPLE OF GAUGE TABLE WITH ZERO LIST AND TRIM

HEIGHTS VOLUMES HEIGHTS VOLUMES HEIGHTS VOLUMES HEIGHTS VOLUMES HEIGHTS VOLUMES

m m3 m m3 m m3 m m3 m m3

0.00 5.51 0.25 233.02 0.50 464.94 0.75 701.27 1.00 942.01

0.01 14.52 0.26 242.21 0.51 474.31 0.76 710.82 1.01 951.73

0.02 23.54 0.27 251.41 0.52 483.69 0.77 720.37 1.02 961.46

0.03 32.57 0.28 260.62 0.53 493.07 0.78 729.93 1.03 971.20

0.04 41.61 0.29 269.83 0.54 502.46 0.79 739.50 1.04 980.94

0.05 50.66 0.30 279.05 0.55 511.86 0.80 749.07 1.05 990.69

0.06 59.71 0.31 288.28 0.56 521.26 0.81 758.65 1.06 1000.45

0.07 68.77 0.32 297.52 0.57 530.67 0.82 768.24 1.07 1010.21

0.08 77.83 0.33 306.76 0.58 540.09 0.83 777.83 1.08 1019.98

0.09 86.90 0.34 316.01 0.59 549.52 0.84 787.43 1.09 1029.76

0.10 95.98 0.35 325.26 0.60 558.95 0.85 797.04 1.10 1039.54

0.11 105.07 0.36 334.52 0.61 568.39 0.86 806.66 1.11 1049.33

0.12 114.16 0.37 343.79 0.62 577.83 0.87 816.28 1.12 1059.13

0.13 123.26 0.38 353.07 0.63 587.29 0.88 825.91 1.13 1068.94

0.14 132.37 0.39 362.36 0.64 596.75 0.89 835.54 1.14 1078.75

0.15 141.49 0.40 371.65 0.65 606.21 0.90 845.19 1.15 1088.57

0.16 150.61 0.41 380.94 0.66 615.69 0.91 854.84 1.16 1098.40

0.17 159.74 0.42 390.25 0.67 625.17 0.92 864.50 1.17 1108.23

0.18 168.87 0.43 399.56 0.68 634.66 0.93 874.16 1.18 1118.07

0.19 178.02 0.44 408.88 0.69 644.15 0.94 883.83 1.19 1127.92

0.20 187.17 0.45 418.21 0.70 653.66 0.95 893.51 1.20 1137.78

0.21 196.32 0.46 427.54 0.71 663.17 0.96 903.20 1.21 1147.64

0.22 205.49 0.47 436.88 0.72 672.68 0.97 912.89 1.22 1157.51

0.23 241.66 0.48 446.23 0.73 682.21 0.98 922.59 1.23 1167.39

0.24 223.84 0.49 455.58 0.74 691.74 0.99 932.30 1.24 1177.27

Totalvolume:25577.680cubicmetres.

68
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

TABLE A4-2: EXAMPLE OF BOTTOM FINE GAUGE TABLE

HEIGHT +0 MM +1 MM +2 MM +3 MM +4 MM +5 MM +6 MM +7 MM +8 MM +9 MM
0.10 41.340 41.750 42.160 42.570 42.980 43.390 43.800 44.210 44.620 45.030
0.11 45.440 45.851 46.262 46.673 47.084 47.495 47.906 48.317 48.728 49.139
0.12 49.550 49.962 50.374 50.786 51.198 51.610 52.022 52.434 52.846 53.258
0.13 53.670 51.082 54.494 54.906 55.318 55.730 56.142 56.554 56.966 57.378
0.14 57.790 58.202 58.614 59.026 59.438 59.850 60.262 60.674 61.086 61.498
0.15 61.910 62.323 62.736 63.149 63.562 63.975 64.388 64.801 65.214 65.627
0.16 66.040 66.454 66.868 67.282 67.696 68.110 68.524 68.938 69.352 69.766
0.17 70.180 70.594 71.008 71.422 71.836 72.250 72.664 73.078 73.492 73.906
0.18 74.320 71.734 75.148 75.562 75.976 76.390 76.804 77.218 77.632 78.046
0.19 78.460 78.875 79.290 79.705 80.120 80.535 80.950 81.365 81.780 82.195
0.20 82.610 83.026 83.442 83.858 84.274 84.690 85.106 85.522 85.938 86.354
0.21 86.770 87.186 87.602 88.018 88.434 88.850 89.266 89.682 90.098 90.514
0.22 90.930 91.347 91.764 92.181 92.598 93.015 93.432 93.849 94.266 94.683
0.23 95.100 95.517 95.934 96.351 96.768 98.185 97.602 98.019 98.436 98.853
0.24 99.270 99.688 100.106 100.524 100.942 101.360 101.778 102.196 102.614 103.032
0.25 103.450 103.868 104.286 104.704 105.122 105.540 105.958 106.376 106.794 107.212
0.26 107.630 108.049 108.468 108.887 109.306 109.725 110.144 110.563 110.982 111.401
0.27 111.820 112.239 112.658 113.077 113.496 113.915 114.334 114.753 115.172 115.591
0.28 116.010 116.430 116.850 117.270 117.690 118.110 118.530 118.950 119.370 119.790
0.29 120.210 120.630 121.050 121.470 121.890 122.310 122.730 123.150 123.570 123.990
0.30 124.410 124.831 125.252 125.673 126.094 126.515 126.936 127.357 127.778 128.199
0.31 128.620 129.042 129.464 129.886 130.308 130.730 131.152 131.574 131.996 132.418
0.32 132.840 133.262 133.684 134.106 134.528 134.950 135.372 135.794 136.216 136.638
0.33 137.060 137.482 137.904 138.326 138.748 139.170 139.592 140.014 140.436 140.858
0.34 141.280 141.703 142.126 142.549 142.972 143.395 143.818 144.241 144.664 145.087
0.35 145.510 145.934 146.358 146.782 147.206 147.630 148.054 148.478 148.902 149.326
0.36 149.750 150.174 150.598 151.022 151.446 151.870 152.294 152.718 153.142 153.566
0.37 153.990 154.414 154.838 155.262 155.686 156.110 156.534 156.958 157.382 157.806
0.38 158.230 158.655 159.080 159.505 159.930 160.355 160.780 161.205 161.630 162.055
0.39 162.480 162.906 163.332 163.758 164.184 164.610 165.036 165.462 165.888 166.314
0.40 166.740 167.166 167.592 168.018 168.444 168.870 169.296 169.722 170.148 170.574
0.41 171.000 171.427 171.854 172.281 172.708 173.135 173.562 173.989 174.416 174.843
0.42 175.270 175.697 176.124 176.551 176.978 177.405 177.832 178.259 178.686 179.113
0.43 179.540 179.968 180.396 180.824 181.252 181.680 182.108 182.536 182.964 183.392
0.44 183.820 184.248 184.676 185.104 185.532 185.960 186.388 186.816 187.244 187.672
0.45 188.100 188.529 188.958 189.387 189.816 190.245 190.674 191.103 191.532 191.961
0.46 192.390 192.819 193.248 193.677 194.106 194.535 194.964 195.393 195.822 196.251
0.47 196.680 197.110 197.540 197.970 198.400 198.830 199.260 199.690 200.120 200.550
0.48 200.980 201.410 201.840 202.270 202.700 203.130 203.560 203.990 204.420 204.850
0.49 205.280 205.711 206.142 206.573 207.004 207.435 207.866 208.297 208.728 209.159
0.50 209.590 210.022 210.454 210.886 211.318 211.750 212.182 212.614 213.046 213.478
0.51 213.910 214.341 214.772 215.203 215.634 216.065 216.496 216.927 217.358 217.789
0.52 218.220 218.653 219.086 219.519 219.952 220.385 220.818 221.251 221.684 222.117
0.53 222.550 222.983 223.416 223.849 224.282 224.715 225.148 225.581 226.014 226.447
0.54 226.880 227.313 227.746 228.179 228.612 229.045 229.478 229.911 230.344 230.777
0.55 231.210 231.644 232.078 232.512 232.946 233.380 233.814 234.248 234.682 235.116
0.56 235.550 235.985 236.420 236.855 237.290 237.725 238.160 238.595 239.030 239.465
0.57 239.900 240.335 240.770 241.205 241.640 242.075 242.510 242.945 243.380 243.815
0.58 244.250 244.686 245.122 245.558 245.994 246.43 246.866 247.302 247.738 248.174
0.59 248.610 249.046 249.482 249.918 250.354 250.790 251.226 251.662 252.098 252.534

Heightsareinmetres,volumesincubicmetres.

69
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

TABLE A4-3: EXAMPLE OF TOP FINE GAUGE TABLE

HEIGHT +0 MM +1 MM +2 MM +3 MM +4 MM +5 MM +6 MM +7 MM +8 MM +9 MM
21.75 26021.430 26022.439 26023.448 26024.457 26025.466 26026.475 26027.484 26028.493 26029.502 26030.511
21.76 26031.520 26032.527 26033.534 26034.541 26035.548 26036.555 26037.562 26038.569 26039.576 26040.583
21.77 26041.590 26042.598 26043.606 26044.614 26045.622 26046.630 26047.638 26048.646 26049.654 26050.662
21.78 26051.670 26052.675 26053.680 26054.685 26055.690 26056.695 26057.700 26058.705 26059.710 26060.715
21.79 26061.720 26062.726 26063.732 26064.738 26065.744 26066.750 26067.756 26068.762 26069.768 26070.774
21.80 26071.780 26072.786 26073.792 26074.798 26075.804 26076.810 26077.816 26078.822 26079.828 26080.834
21.81 26081.840 26082.843 26083.846 26084.849 26085.852 26086.855 26087.858 26088.861 26089.864 26090.867
21.82 26091.870 26092.873 26093.876 26094.879 26095.882 29096.885 26097.888 26098.891 26099.894 26100.897
21.83 26101.900 26102.903 26103.906 26104.909 26105.912 26106.915 26107.918 26108.921 26109.924 26110.927
21.84 26111.930 26112.932 26113.934 26114.936 26115.938 26116.940 26117.942 26118.944 26119.946 26120.948
21.85 26121.950 26122.951 26123.952 26124.953 26125.954 26126.955 26127.956 26128.957 26129.958 26130.959
21.86 26131.960 26132.961 26133.962 26134.963 26135.964 26136.965 26137.966 26138.967 26139.968 26140.969
21.87 26141.970 26142.969 26143.968 26144.967 26145.966 26146.965 26147.964 26148.963 26149.962 26150.961
21.88 26151.960 26152.959 26153.958 26154.957 26155.956 26156.955 26157.954 26158.953 26159.952 26160.951
21.89 26161.950 26162.950 26163.950 26164.950 26165.950 26166.950 26167.950 26168.950 26169.950 26170.950
21.90 26171.950 26172.947 26173.944 26174.941 26175.938 26176.935 26177.932 26178.929 26179.926 26180.923
21.91 26181.920 26182.918 26183.916 26184.914 26185.912 26186.910 26187.908 26188.906 26189.904 26190.902
21.92 26191.900 26192.896 26193.892 16194.888 26195.884 26196.880 26197.876 26198.872 26199.868 26200.864
21.93 26201.860 26202.855 26203.850 26204.845 26205.840 26206.835 26207.830 26208.825 26209.820 26210.815
21.94 26211.810 26212.806 26213.802 26214.798 26215.794 26216.790 26217.786 26218.782 26219.778 26220.774
21.95 26221.770 26222.763 26223.756 26224.749 26225.742 26226.735 26227.728 26228.721 26229.714 26230.707
21.96 26231.700 26232.693 26233.686 26234.679 26235.672 26236.665 26237.658 26238.651 26239.644 26240.637
21.97 26241.630 26242.624 26243.618 26244.613 26245.606 26246.600 26247.594 26248.588 26249.582 26250.576
21.98 26251.570 26252.562 26253.554 26254.546 26255.538 26256.530 26257.522 26258.514 26259.506 26260.498
21.99 26261.490 26262.480 26263.470 26264.460 26265.450 26266.440 26267.430 26268.420 26269.410 26270.400
22.00 23271.390 26272.383 26273.376 26274.369 26275.362 26276.355 26277.348 26278.341 26279.334 26280.327
22.01 26281.320 26282.309 26283.298 26284.287 26285.276 26286.265 26287.254 26288.243 26289.232 26290.221
22.02 26291.210 26292.200 26293.190 26294.180 26295.170 26296.160 26297.150 26298.140 26299.130 26300.120
22.03 26301.110 26302.098 26303.086 26304.074 26305.062 26306.050 26307.038 26308.026 26309.014 26310.002
22.04 26310.990 26311.978 26312.966 26313.954 26314.942 26315.930 26316.918 26317.906 26318.894 26319.882
22.05 26320.870 26321.858 26322.846 26323.834 26324.822 26325.810 26326.798 26327.786 26328.774 26329.762
22.06 26330.750 26331.736 26332.722 26333.708 26334.694 26335.680 26336.666 26337.652 26338.638 26339.624
22.07 26340.610 26341.595 26342.580 26343.565 26344.550 26345.535 26346.520 26347.505 26348.490 26349.475
22.08 26350.460 26351.446 26352.432 26353.418 26354.404 26355.390 26356.376 26357.362 26358.348 26359.334
22.09 26360.320 26361.304 26362.288 26363.272 26364.256 26365.240 26366.224 26367.208 26368.192 26369.176
22.10 26370.160 26371.144 26372.128 26373.112 26374.096 26375.080 26376.064 26377.048 26378.032 26379.016
22.11 26380.000 26380.984 26381.968 26382.952 26383.936 26384.920 26385.904 26386.888 26387.872 26388.856
22.12 26389.840 26390.821 26391.802 26392.783 26393.764 26394.745 26395.726 26396.707 26397.688 26398.669
22.13 26399.650 26400.631 26401.612 26402.593 26403.574 26404.555 26405.536 26406.517 26407.498 26408.479
22.14 26409.460 26410.442 26411.424 26412.406 26413.388 26414.370 26415.352 26416.334 26417.316 26418.298
22.15 26419.280 26420.260 26421.240 26422.220 26423.200 26424.180 26425.160 26426.140 26427.120 26428.100
22.16 26429.080 26430.060 26431.040 26432.020 26433.000 26433.980 26434.960 26435.940 26436.920 26437.900
22.17 26438.880 26439.858 26440.836 26441.814 26442.792 26443.770 26444.748 26445.726 26446.704 26447.682
22.18 26448.660 26449.638 26450.616 26451.594 26452.572 26453.550 26454.528 26455.506 26456.484 26457.462
22.19 26458.440 26459.418 26460.396 26461.374 26462.352 26463.330 26464.308 26465.286 26466.264 26467.242
22.20 26468.220 26469.196 26470.172 26471.148 26472.124 26473.100 26474.076 26475.052 26476.028 26477.004
22.21 26477.980 26478.955 26479.930 26480.905 26481.880 26482.855 26483.830 26484.805 26485.780 26486.755
22.22 26487.730 26488.707 26489.684 26490.661 26491.630 26492.615 26493.592 26494.569 26495.546 26496.523
22.23 26497.500 26498.473 26499.446 26500.419 26501.392 26502.365 26503.338 26504.311 26505.284 26506.257

Heightsareinmetres,volumesincubicmetres.

70
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

TABLE A4-4: EXAMPLE OF VOLUME CORRECTIONS ACCORDING TO TANK SERVICE TEMPERATURES


FOR SELF SUPPORTING TANKS

Vt = K . V-160C

T (C) K T (C) K
165.0 0.99980 160.8 0.99997
164.9 0.99980 160.7 0.99997
164.8 0.99981 160.6 0.99998
164.7 0.99981 160.5 0.99998
164.6 0.99981 160.4 0.99998
164.5 0.99982 160.3 0.99999
164.4 0.99982 160.2 0.99999
164.3 0.99983 160.1 1.00000
164.2 0.99983 160.0 1.00000
164.1 0.99983 159.9 1.00000
164.0 0.99984 159.8 1.00001
163.9 0.99984 159.7 1.00001
163.8 0.99985 159.6 1.00002
163.7 0.99985 159.5 1.00002
163.6 0.99985 159.4 1.00002
163.5 0.99986 159.3 1.00003
163.4 0.99986 159.2 1.00003
163.3 0.99987 159.1 1.00004
163.2 0.99987 159.0 1.00004
163.1 0.99987 158.9 1.00005
163.0 0.99988 158.8 1.00005
162.9 0.99988 158.7 1.00005
162.8 0.99989 158.6 1.00006
162.7 0.99989 158.5 1.00006
162.6 0.99989 158.4 1.00007
162.5 0.99990 158.3 1.00007
162.4 0.99990 158.2 1.00008
162.3 0.99991 158.1 1.00008
162.2 0.99991 158.0 1.00008
162.1 0.99991 157.9 1.00009
162.0 0.99992 157.8 1.00009
161.9 0.99992 157.7 1.00010
161.8 0.99993 157.6 1.00010
161.7 0.99993 157.5 1.00010
161.6 0.99993 157.4 1.00011
161.5 0.99994 157.3 1.00011
161.4 0.99994 157.2 1.00012
161.3 0.99995 157.1 1.00012
161.2 0.99995 157.0 1.00013
161.1 0.99995 156.9 1.00013
161.0 0.99996 156.8 1.00013
160.9 0.99996 156.7 1.00014

71
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

APPENDIX 5: EXAMPLE OF VOLUME CALCULATION (see Section 3.3)

TABLE A5-1: VOLUME DETERMINATION BEFORE LOADING LNG


LNG CARRIER CARGO ON BOARD AT ARRIVAL
VOYAGENo. SURVEYDATE:
3
CARGODENSITY:451.61kg/m SURVEYTIMEGMT:
TRIM:+80cm(AFT) LOCALTIME:
LISTANGLE:0.5(PORT) REMARKS:
BERTH:


Corrections
Vapour Level Corrected Liquid
temp. gauge height Volume
C reading Shrink. Density List Trim Overall mm m
3

(1) mm mm mm mm mm mm (7) (8)


(2) (3) (4) (5) (6)

TANK1 117.22 575 +42 +2 +4 30 +18 593 249.918

TANK2 124.32 435 +44 +2 +4 51 1 434 425.150

TANK3 125.16 458 +44 +2 +4 56 6 452 509.560

TANK4 124.63 505 +44 +2 +4 56 6 499 563.084

TANK5 122.46 543 +44 +2 +4 54 4 539 502.460

TOTALO/B 2250.172
Notes:

(1) For measurement of temperature in the gaseous phase, see Section 4.2.

(2) Correction for ribbon shrinkage of the float level gauge due to the cryogenic temperature in the gaseous phase
according to Table A5-3. The correction shrink is generally referred to as tape correction. On Moss ships,
there is also a correction called shrinkage factor.

(3) Correction for LNG density, established from the density calculated on the basis of the LNG composition,
according to Table A5-4.

(4) Correction for the list corresponding to the liquid height according to the correction tables, an example of
which is shown in Table A5-5 for tank No. 1. In this case the position of the gauge is at the starboard side of
the ship's centre line.

(5) Correction for the trim corresponding to the liquid height according to the correction tables, an example of
which is shown in Table A5-6 for tank No. 1 (an interpolation was made in this case between the correction
values for a 50 cm and a 100 cm trim). In this case the position of the gauge is on the stern-side of the sideline
of the tank.

(6) Algebraic sum of corrections (2), (3), (4) and (5).

(7) Corrected height resulting from the algebraic sum of the five previous columns.

(8) Determination of the liquid volume given in the calibration tables from the corrected height. These tables are
established for heights varying mm by mm, see example in Table A4-1 (Appendix4) on the basis of certified
tables indicating volumes for heights varying cm by cm. An example is given in Table A4-2 (Appendix 4) for
the empty tank No. 1 and in Table A4-3 (Appendix4) for the full tank No. 2.

72
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

TABLE A5-2: VOLUME DETERMINATION AFTER LOADING LNG

LNG CARRIER CARGO ON BOARD AT DEPARTURE


VOYAGENo. SURVEYDATE:
3
CARGODENSITY:450.90kg/m SURVEYTIMEGMT:
TRIM:+50cm(AFT) LOCALTIME:
LISTANGLE:0 REMARKS:
BERTH:


Corrections
Vapour Level Corrected Liquid
temp. gauge height Volume
C reading Shrink. Density List Trim Overall mm m3
(1) mm mm mm mm mm mm (7) (8)
(2) (3) (4) (5) (6)

TANK1 152.00 22040 1 2 0 19 16 22024 12833.320

TANK2 152.00 22123 1 2 0 32 29 22094 26364.256

TANK3 152.00 22138 1 2 0 35 32 22106 29914.104

TANK4 156.00 22165 1 2 0 35 32 22133 29962.222

TANK5 156.00 22150 1 2 0 34 31 22119 25210.599

TOTALO/Bdep. 124284.501
TOTALO/Barr. 2250.172
%Cargoonboard:98 TOTALloaded 122034.329

Notes(1)to(8):seetableabove.

73
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

TABLE A5-3: EXAMPLE OF GAUGE CORRECTIONS FOR LOW TEMPERATURES

VAPOUR TEMP -165C -160C -155C -150C -145C -140C -135C -130C -125C -120C -115C -110C
METRES mm mm mm mm mm mm mm mm mm mm mm mm

0.250 56 55 54 52 51 49 48 46 45 43 42 40
0.500 56 54 53 52 50 49 47 46 44 43 42 40
0.750 55 54 52 51 50 48 47 45 44 43 41 40
1.000 55 53 52 50 49 48 46 45 43 42 41 39
1.250 54 53 51 50 49 47 46 44 43 42 40 39
1.500 53 52 51 49 48 47 45 44 42 41 40 38
1.750 53 51 50 49 47 46 45 43 42 41 39 38
2.000 52 51 49 48 47 46 44 43 41 40 39 37
2.250 51 50 49 48 46 45 44 42 41 40 38 37
2.500 51 50 48 47 46 44 43 42 41 39 38 36
2.750 50 49 48 46 45 44 43 41 40 39 37 36
3.000 50 48 47 46 45 43 42 41 40 38 37 36
3.250 49 48 47 45 44 43 42 40 39 38 36 35
3.500 48 47 46 45 44 42 41 40 39 37 36 35
3.750 48 47 45 44 43 42 41 39 38 37 36 34
4.000 47 46 45 44 42 41 40 39 38 36 35 34
4.250 47 45 44 43 42 41 39 38 37 36 35 33
4.500 46 45 44 42 41 40 39 38 37 35 34 33
4.750 45 44 43 42 41 40 38 37 36 35 34 32
5.000 45 44 42 41 40 39 38 37 36 34 33 32
5.250 44 43 42 41 40 38 37 36 35 34 33 32
5.500 43 42 41 40 39 38 37 36 35 33 32 31
5.750 43 42 41 40 38 37 36 35 34 33 32 31
6.000 42 41 40 39 38 37 36 35 34 32 31 30
6.250 42 41 39 38 37 36 35 34 33 32 31 30
6.500 41 40 39 38 37 36 35 34 33 32 30 29
6.750 40 39 38 37 36 35 34 33 32 31 30 29
7.000 40 39 38 37 36 35 34 33 32 31 30 28
7.250 39 38 37 36 35 34 33 32 31 30 29 28
7.500 38 37 37 36 35 34 33 32 31 30 29 28
7.750 38 37 36 35 34 33 32 31 30 29 28 27
8.000 37 36 35 34 33 33 32 31 30 29 28 27
8.250 37 36 35 34 33 32 31 30 29 28 27 26
8.500 36 35 34 33 32 31 31 30 29 28 27 26
8.750 35 34 34 33 32 31 30 29 28 27 26 25
9.000 35 34 33 32 31 30 29 29 28 27 26 25
9.250 34 33 32 32 31 30 29 28 27 26 25 24
9.500 33 33 32 31 30 29 28 28 27 26 25 24
9.750 33 32 31 30 30 29 28 27 26 25 24 24
10.000 32 31 31 30 29 28 27 27 26 25 24 23
10.250 32 31 30 29 28 28 27 26 25 24 24 23
10.500 31 30 29 29 28 27 26 26 25 24 23 22
10.750 30 30 29 28 27 27 26 25 24 23 23 22
11.000 30 29 28 28 27 26 25 24 24 23 22 21
11.250 29 28 28 27 26 25 25 24 23 22 22 21
11.500 29 28 27 26 26 25 24 23 23 22 21 20
11.750 28 27 27 26 25 24 24 23 22 22 21 20
12.000 27 27 26 25 25 24 23 22 22 21 20 20
12.250 27 26 25 25 24 23 23 22 21 21 20 19
12.500 26 25 25 24 23 23 22 21 21 20 19 19

74
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

TABLE A5-4: FLOAT GAUGES EXAMPLE OF FLOAT BUOYANCY / DIP CORRECTIONS


ThefloatgaugeshavebeenadjustedtoreadcorrectlywiththefloatfloatinginLNGofadensityof470kg/m3.

Thefollowingtablegivescorrectionstoapplytothefloatgaugereadingswhenfloatinginliquidsofdensitiesother
3
than470kg/m .

DENSITY CORRECTIONS
3
(kg/m ) (mm)
450452 +2
453462 +1
463472 0
473482 1
483493 2
494505 3
506516 4
517529 5
530542 6
543550 7

TABLE A5-5: EXAMPLE OF LIST CORRECTION TABLE

LIST ANGLE ()

HEIGHTS -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.5 1.0 1.5 2.0 2.5 3.0
0.51 0.034 0.026 0.019 0.013 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.008 -0.006
0.52 0.034 0.026 0.019 0.013 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.008 -0.007
0.53 0.033 0.026 0.019 0.013 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.008 -0.007
0.57 0.033 0.026 0.019 0.013 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.008 -0.007
0.58 0.033 0.026 0.019 0.013 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.008 -0.008
0.81 0.033 0.026 0.019 0.013 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.008 -0.008
0.82 0.033 0.026 0.019 0.013 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.008 -0.007
0.88 0.033 0.026 0.019 0.013 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.008 -0.007
0.89 0.034 0.026 0.019 0.013 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.008 -0.007
1.01 0.034 0.026 0.019 0.013 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.008 -0.007
1.02 0.034 0.026 0.019 0.014 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.008 -0.007
1.05 0.034 0.026 0.019 0.014 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.008 -0.007
1.06 0.034 0.026 0.020 0.014 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.008 -0.007
1.36 0.034 0.026 0.020 0.014 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.008 -0.007
1.37 0.034 0.027 0.020 0.014 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.008 -0.007
1.43 0.034 0.027 0.020 0.014 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.008 -0.007
1.44 0.034 0.027 0.020 0.014 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.007 -0.007
1.47 0.034 0.027 0.020 0.014 0.008 0.004 -0.003 -0.005 -0.007 -0.008 -0.007 -0.007
1.48 0.034 0.027 0.020 0.014 0.008 0.004 -0.003 -0.005 -0.007 -0.007 -0.007 -0.007
1.59 0.034 0.027 0.020 0.014 0.008 0.004 -0.003 -0.005 -0.007 -0.007 -0.007 -0.007
1.60 0.034 0.027 0.020 0.014 0.008 0.004 -0.003 -0.005 -0.007 -0.007 -0.007 -0.006
1.68 0.034 0.027 0.020 0.014 0.008 0.004 -0.003 -0.005 -0.007 -0.007 -0.007 -0.006
1.69 0.035 0.027 0.020 0.014 0.008 0.004 -0.003 -0.005 -0.007 -0.007 -0.007 -0.006
2.29 0.035 0.027 0.020 0.014 0.008 0.004 -0.003 -0.005 -0.007 -0.007 -0.007 -0.006
2.30 0.035 0.027 0.020 0.014 0.008 0.004 -0.003 -0.005 -0.006 -0.007 -0.007 -0.006

Thelistanglesareindegrees,negativevaluestoport,andpositivevaluestostarboard.
Heightsandcorrectionsareinmetres.

75
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

TABLE A5-6: EXAMPLE OF TRIM CORRECTION TABLE

TRIM

HEIGHTS -2.00 -1.50 -1.00 -0.50 0.50 1.00 1.50

0.55 0.077 0.057 0.038 0.019 -0.019 -0.038 -0.057


0.56 0.077 0.057 0.038 0.019 -0.019 -0.038 -0.057
0.59 0.077 0.057 0.038 0.019 -0.019 -0.038 -0.057
0.60 0.077 0.058 0.038 0.019 -0.019 -0.038 -0.057
0.61 0.077 0.058 0.038 0.019 -0.019 -0.038 -0.057
0.62 0.077 0.058 0.038 0.019 -0.019 -0.038 -0.057
0.67 0.077 0.058 0.038 0.019 -0.019 -0.038 -0.057
0.68 0.077 0.058 0.038 0.019 -0.019 -0.038 -0.057
0.72 0.077 0.058 0.038 0.019 -0.019 -0.038 -0.057
0.73 0.077 0.058 0.038 0.019 -0.019 -0.038 -0.057
0.78 0.077 0.058 0.038 0.019 -0.019 -0.038 -0.057
0.79 0.077 0.058 0.038 0.019 -0.019 -0.038 -0.057
0.80 0.077 0.058 0.038 0.019 -0.019 -0.038 -0.057
0.81 0.077 0.058 0.039 0.019 -0.019 -0.038 -0.057
0.82 0.077 0.058 0.039 0.019 -0.019 -0.038 -0.057
0.83 0.077 0.058 0.039 0.019 -0.019 -0.038 -0.057
0.84 0.077 0.058 0.039 0.019 -0.019 -0.038 -0.057
0.85 0.077 0.058 0.039 0.019 -0.019 -0.038 -0.057
0.86 0.077 0.058 0.039 0.019 -0.019 -0.038 -0.057
0.87 0.077 0.058 0.039 0.019 -0.019 -0.038 -0.058
0.89 0.077 0.058 0.039 0.019 -0.019 -0.038 -0.058
0.90 0.077 0.058 0.039 0.019 -0.019 -0.038 -0.058
0.92 0.077 0.058 0.039 0.019 -0.019 -0.038 -0.058
0.93 0.077 0.058 0.039 0.019 -0.019 -0.038 -0.058
0.94 0.077 0.058 0.039 0.019 -0.019 -0.039 -0.058

Thetrimvaluesareinmetres,negativevaluesfore,positivevaluesaft.
Theheightsandthecorrectionsareinmetres.

76
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

APPENDIX 6: EXAMPLES OF DISCONTINUOUS AND CONTINUOUS SAMPLING AND VAPORIZATION


SYSTEMS

a) LNGsamplingsystemwithelectricLNGvaporizer

77
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

b) LNGsamplingsystemwithsteamvaporizer




c) Continuoussamplingsystemwithcompressorforcharginggassamplecontainers

78
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

d) Continuoussamplingsystemwithpressurizedargongasforemptyingsamplingvesselandfillinggassample
containers

79
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

APPENDIX 7: DIRECT IN-LINE ANALYSIS WITH RAMAN SPECTROSCOPY

New technology, for information only, see reference [11]

The Technology

Raman spectroscopy is a form of vibration spectroscopy where a material of interest is illuminated by a highly
monochromaticlightsource,typicallyalaser,andtheresultingscatteredlightisanalyzed.Whenlightinteractswith
amolecule,mostphotonsareelasticallyscattered.Thescatteredphotonshavethesameenergy(frequency)and,
8
therefore wavelength (colour) as the incident photons. However, a small fraction of light (approximately 1 in 10
photons) is scattered at optical frequencies different from, and usually lower than, the frequency of the incident
photons. The process leading to this inelastic scatter is termed the Raman effect, discovered in 1928 by Indian
physicistChandrasekharaVenkataRamanwhowasawardedtheNobelPrizeforphysicsin1930.

Theenergyofamolecularvibrationmodedependsonthemolecularstructureofthematerial.ARamanspectrum
is,ineffect,amolecularfingerprintuniquetothatcompound(seeFigureA7.1).Thespectrumcanbeanalyzedto
obtain information on the molecular components of the material being illuminated as well as their relative
abundance. The signal strength is strongly dependent on the density of the sample so it is advantageous to
measuretheliquiddirectly.Thisalsoavoidstheneedforthephasechangeormolecularseparationcommonwith
other analytical methods. Due to the fact that the Raman effect is a relatively weak process, with the currently
availabletechnologyitistypicallynotsuitablefortrace(lowppm)analysisofhigher(C7+)componentsinLNG.

The Adaptation to LNG

ForLNGanalysisacryogenicopticalprobeandaninternalcalibrationprocessweredeveloped.Thelatterallows
theanalysertoperformdiagnosticsautomatically.

ViaafibreopticcableandtheopticalprobethelaserlightisintroducedintotheLNGpipelinethroughasapphire
windowattheendoftheprobe.Theactualmolecularexcitationtakesplaceapproximately300to400micronsoff
thewindow.Asthereisnowet chemistrytakingplacethereisno need for cleaningthedevice orotherregular
maintenanceissues.Theprobecontainsaseriesofopticalcomponentsforfilteringthelightandisconnectedtoa
baseunitbyalengthoffibreopticcable.

Thefibreopticcableisterminatedintheanalyserbaseunitwhichcanbeinstalledinthefieldorinsideatechnical
building, e.g. a process control room. An initial calibration of the equipment is performed at the factory prior to
shipment.Onceinstalledonsite,asimplecalibrationofthefibretransmissionintensityisallthecalibrationthatis
needed.Analysermaintenancetypicallyoccursonanannualbasis.

Laboratory and LNG field testing allowed for the development of a mathematical model to convert the Raman
spectrum to a composition and energy content. This model will be enhanced as additional field data is gathered,
thus eliminatingthe variationthatresultsfromindividualcalibrations ofcurrentcomposition analysis systemsthat
arerequiredbygaschromatographs.

Field Testing

DatawascollectedonseveralLNGcarrierunloadingoperations.Theresultsofthemeasurementsindicatedthat2
sigma repeatability is better than 0.1 BTU and accuracy is better than 0.5 BTU when compared to gas
chromatographic data taken at the same time (see Figure A7.2). After analysis, the bulk of the 0.5 BTU offset
appears to be due to the BTU value assigned by the chromatograph to C6+ components, which are not (yet)
incorporatedinthecurrentRamanModel.

Futuremodelrefinements,whichwillincludetheeffectsoftheC6+componentsshouldbeabletoclosetheoffset.
Thismaybefurtherimprovedthroughtheuseofamoresophisticatedmodelthatincorporatesmoreextensivefield
andlaboratoryanalysis.

80
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

FigureA7.1GraphicalrepresentationoftheRamaneffect.



FigureA7.2GCandRamanBTUdatacomparisonforashipunloading.

81
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

APPENDIX 8: Grubbs test [19] and [20]

GrubbstestisspecifiedinISO57252standard,asfollows:

One outlying observation

Givenasetofdataxifori=1,2,,p,arrangedinascendingorder,todeterminewhetherthelargestobservationis
anoutlierusingGrubbstest,computetheGrubbsstatistic,Gp:
= /
where
1
=

and

1
= ( )
1

Totestthesignificanceofthesmallestobservation,computetheteststatisticG1:
= ( )/
a) Iftheteststatisticislessthanorequaltoits5%criticalvalue,theitemtestedisacceptedascorrect.
b) Iftheteststatisticisgreaterthanits5%criticalvalueandlessthanorequaltoits1%criticalvalue,theitem
testediscalledastragglerandisindicatedbyasingleasterisk.
c) Iftheteststatisticisgreaterthanits1%criticalvalue,theitemiscalledastatisticaloutlierandisindicatedby
adoubleasterisk.

Two outlying observations

Totestwhetherthetwolargestobservationsmaybeoutliers,computetheGrubbsteststatisticG:

= , /
where

= ( )

and

, = ,

and

1
, =
2

Alternatively,totestthetwosmallestobservations,computetheGrubbsteststatisticG:

= , /
where

, = ,

and
1
, =
2

CriticalvaluesfortheGrubbstestaregivenintable5ofISO57252(seetableA81).

Application of Grubbs test

Whenanalysingaprecisionexperiment,Grubbstestcanbeappliedasfollows:
ApplytheGrubbstestforoneoutlyingobservation.Ifitisshowntobeanoutlier,excludeit,andrepeatthetestat
the other extreme (e.g. if the highest is an outlier then look at the lowest with the highest excluded), but do not
apply the Grubbs test for two outlying observations. If the Grubbs test does not show any observation to be an
outlier,thenapplythedoubleGrubbstest.

82
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

Table A8-1: Critical values for Grubbs test (ISO 5725-2)

Note:Thistableisextendedindocument[20]top=147.

83
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

Numerical example

Below, the Grubbs test is applied to an example of GC analyses of vaporised LNG, in order to determine the
presence of outliers, and eliminate them before calculating the average LNG composition. Table A82 shows the
datafromtheGC.

Notethatifonecomponentisdetectedasoutlier,thewholeanalysismustbeeliminated.

Table A8-2: Example GC analyses

Analysis CH4 C2H6 C3H8 iC4 nC4 iC5 nC5 nC6 N2 CO2
1 95,653 3,423 0,728 0,095 0,084 0,002 0,001 0,001 0,014 0,000
2 95,877 3,289 0,603 0,099 0,102 0,005 0,002 0,001 0,024 0,000
3 95,657 3,421 0,726 0,096 0,083 0,002 0,001 0,000 0,014 0,000
4 95,588 3,449 0,751 0,103 0,091 0,003 0,001 0,000 0,014 0,000
5 95,500 3,504 0,769 0,111 0,099 0,003 0,001 0,001 0,013 0,000
6 95,710 3,386 0,714 0,094 0,081 0,002 0,000 0,000 0,013 0,000
7 95,671 3,417 0,719 0,095 0,083 0,002 0,001 0,000 0,014 0,000
8 95,494 3,446 0,774 0,131 0,133 0,006 0,003 0,001 0,013 0,000
9 95,626 3,439 0,734 0,098 0,086 0,002 0,001 0,001 0,013 0,000
10 95,864 3,346 0,596 0,087 0,087 0,004 0,002 0,001 0,013 0,000
11 95,584 3,452 0,747 0,106 0,096 0,003 0,000 0,000 0,012 0,000
12 95,545 3,501 0,752 0,100 0,087 0,003 0,000 0,000 0,013 0,000

Firstly,theGrubbsstatisticGpiscalculatedforeachLNGcomponentandcomparedtothecriticalvalues(seetable
A83).ThevaluesofGcriticalareobtainedfromtableA81,forp=12.

Table A8-3: Example Gp statistic

CH4 C2H6 C3H8 iC4 nC4 iC5 nC5 nC6 N2 CO2


xp 95,877 3,504 0,774 0,131 0,133 0,006 0,003 0,001 0,024 0,000
Gp 1,858 1,343 0,965 2,652 2,805 2,224 2,129 0,957 3,114 0,000
Gcritical 5% 2,412
Gcritical 1% 2,636
GpOk? YES YES YES ** ** YES YES YES ** YES

AsitisshownintableA83,thehighestvaluesforthecomponentsiC4,nC4andN2havebeendetectedasoutliers;
thus,thetwoanalysescontainingthesevaluesareexcluded(analyses2and8).Next,thetestisrepeatedtothe10
analysesremaining,computingtheG1statistic.ThevaluesofGcriticalareobtainedfromtableA81,forp=10.

Table A8-4: Example G1 statistic

CH4 C2H6 C3H8 iC4 nC4 iC5 nC5 nC6 N2 CO2


x1 95,500 3,346 0,596 0,087 0,081 0,002 0,000 0,000 0,012 0,000
G1 1,389 1,836 2,661 1,686 1,134 0,858 1,265 0,775 1,926 0,000
Gcritical 5% 2,290
Gcritical 1% 2,482
G1Ok? YES YES ** YES YES YES YES YES YES YES

Thelowestvalueof propaneisdetectedasoutlier,andtheanalysis10whereitiscontainedisexcluded.Finally,
theaveragecompositioniscalculatedwiththe9analysesthathavebeenacceptedascorrectbytheGrubbstest.

84
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

Table A8-5: Example Average composition

Analysis CH4 C2H6 C3H8 iC4 nC4 iC5 nC5 nC6 N2 CO2
1 95,653 3,423 0,728 0,095 0,084 0,002 0,001 0,001 0,014 0,000
3 95,657 3,421 0,726 0,096 0,083 0,002 0,001 0,000 0,014 0,000
4 95,588 3,449 0,751 0,103 0,091 0,003 0,001 0,000 0,014 0,000
5 95,500 3,504 0,769 0,111 0,099 0,003 0,001 0,001 0,013 0,000
6 95,710 3,386 0,714 0,094 0,081 0,002 0,000 0,000 0,013 0,000
7 95,671 3,417 0,719 0,095 0,083 0,002 0,001 0,000 0,014 0,000
9 95,626 3,439 0,734 0,098 0,086 0,002 0,001 0,001 0,013 0,000
11 95,584 3,452 0,747 0,106 0,096 0,003 0,000 0,000 0,012 0,000
12 95,545 3,501 0,752 0,100 0,087 0,003 0,000 0,000 0,013 0,000
95,615 3,444 0,738 0,100 0,088 0,002 0,001 0,000 0,013 0,000

85
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

APPENDIX 9: TABLES FOR LNG DENSITY CALCULATION ACCORDING TO NBS (see Section 9)
TABLE A9-1: COMPONENT MOLAR VOLUMES

Molar volume, l/mol


Component
118 K 116 K 114 K 112 K 110 K 108 K 106 K

CH4 0.038817 0.038536 0.038262 0.037995 0.037735 0.037481 0.037234

C2H6 0.048356 0.048184 0.048014 0.047845 0.047678 0.047512 0.047348

C3H8 0.062939 0.062756 0.062574 0.062392 0.062212 0.062033 0.061855

iC4H10 0.078844 0.078640 0.078438 0.078236 0.078035 0.077836 0.077637

nC4H10 0.077344 0.077150 0.076957 0.076765 0.076574 0.076384 0.076194

iC5H12 0.092251 0.092032 0.091814 0.091596 0.091379 0.091163 0.090948

nC5H12 0.092095 0.091884 0.091673 0.091462 0.091252 0.091042 0.090833

N2 0.050885 0.049179 0.047602 0.046231 0.045031 0.043963 0.043002

Source:N.B.S.Technicalnote1030December1980.

3
TABLE A9-2: VOLUME CORRECTION FACTOR, K1 . 10

Molecular Volume reduction, l/mol


weight
of mixture
g/mol 105 K 110 K 115 K 120 K 125 K 130 K 135 K

16 0.007 0.008 0.009 0.010 0.013 0.015 0.017

17 0.165 0.180 0.220 0.250 0.295 0.345 0.400

18 0.340 0.375 0.440 0.500 0.590 0.700 0.825

19 0.475 0.535 0.610 0.695 0.795 0.920 1.060

20 0.635 0.725 0.810 0.920 1.035 1.200 1.390

21 0.735 0.835 0.945 1.055 1.210 1.370 1.590

22 0.840 0.950 1.065 1.205 1.385 1.555 1.800

23 0.920 1.055 1.180 1.330 1.525 1.715 1.950

24 1.045 1.155 1.280 1.450 1.640 1.860 2.105

25 1.120 1.245 1.380 1.550 1.750 1.990 2.272

Source:N.B.S.Technicalnote1030December1980.

86
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted
3
TABLE A9-3: VOLUME CORRECTION FACTOR, K2 . 10

Molecular Volume reduction, l/mol


weight
of mixture 105 K 110 K 115 K 120 K 125 K 130 K 135 K

16 0.010 0.015 0.024 0.032 0.043 0.058 0.075

17 0.240 0.320 0.410 0.600 0.710 0.950 1.300

18 0.420 0.590 0.720 0.910 1.130 1.460 2.000

19 0.610 0.770 0.950 1.230 1.480 1.920 2.400

20 0.750 0.920 1.150 1.430 1.730 2.200 2.600

21 0.910 1.070 1.220 1.630 1.980 2.420 3.000

22 1.050 1.220 1.300 1.850 2.230 2.680 3.400

23 1.190 1.370 1.450 2.080 2.480 3.000 3.770

24 1.330 1.520 1.650 2.300 2.750 3.320 3.990

25 1.450 1.710 2.000 2.450 2.900 3.520 4.230

Source:N.B.S.Technicalnote1030December1980.

87
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

APPENDIX 10: EXAMPLE OF LNG DENSITY CALCULATION (see Section 9)

ForLNGwithcompositionasindicatedinTableA101andA102andatatemperatureof112.38K

TABLE A10-1

1) CALCULATIONOFMOLECULARWEIGHTINg/mol(inaccordancewithISO6976:1995)
Molar composition Molecular weight
Component Molecular
Molar fraction
Component molecular weight weight fract.
mol/mol
g/mol g/mol
MethaneCH4 0.89235 16.043 14.315971

EthaneC2H6 0.08267 30.070 2.485887

PropaneC3H8 0.01313 44.097 0.578994

IsobutaneIc4H10 0.00167 58.123 0.097065

nButanenC4H10 0.00277 58.123 0.161001

IsopentaneiC5H12 0.00011 72.150 0.007937

nPentanenC5H12 0.00000 72.150 0.000000

nHexanenC6H14 0.00000 86.177 0.000000

NitrogenN2 0.00730 28.0135 0.204499

TOTAL 1.00000 17.851354

TABLE A10-2

2) CALCULATIONOFTHEMOLARVOLUMEINl/molBYTEMPERATUREINTERPOLATION

AtT=112.38KMolarvolume=icomponentmolarvolumes=i(molefractionsxmolarvolumes@112.38K)
molarvolumes@112.38K=linearinterpolationbetweenmolarvolume@112Kand114K.

112.38 112
= .( 0.038262 0.037995 ) 0.037995 0.038046
114 112
Differential Component
Molar volume from Differential Molar
Mole molar molar
Component Table No. 14 molar volume volume at
fraction volume for volume at
for 1.62 K 112.38 K
at 114 K at 112 K 2K 112.38 K
(3) (4) (5)=(3)(4) (6)=(1.62K/2K).(5) (7)=(3)(6) (8)=(7).(2)
(1) (2)
l/mol l/mol l/mol l/mol l/mol l/mol

CH4 0.892350 0.038262 0.037995 0.000267 0.000216 0.038046 0.033950

C2H6 0.082670 0.048014 0.047845 0.000169 0.000137 0.047877 0.003958

C3H8 0.013130 0.062574 0.062392 0.000182 0.000147 0.062427 0.000820

iC4H10 0.001670 0.078438 0.078236 0.000202 0.000164 0.078274 0.000131

nC4H10 0.002770 0.076957 0.076765 0.000192 0.000156 0.076801 0.000213

iC5H12 0.000110 0.091814 0.091596 0.000218 0.000171 0.091637 0.000010

N2 0.007300 0.047602 0.046231 0.001371 0.001111 0.046491 0.000339

TOTAL 1.000000 0.039421

88
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

3)CALCULATIONOFVOLUMEREDUCTIONFACTORSk1ANDk2

3.1) Calculationofk1

Thek1valuesaregivenfordifferenttemperaturesandmolecularweightsinTableA92.

Twointerpolationsmustbecarriedout:
1) onthetemperature,
2) onthemolecularweight.

TABLE A10-3

T=112.38K

CALCULATION OF THE CORRECTION FACTOR k1 in l/mol - temperature interpolation



Differential
Molecular k1 at k1 at Differential value k1 at
value for k1 for
weight 115 K T1 110 K T2 for 2.62 K 112.38 K
5K
(11) (12) (13)=(11)(12) (14)=(2.62K/5K).(13) (15)=(11)(14)
g/mol
l/mol l/mol l/mol l/mol l/mol

MW2(17g/mol) 0.000220 0.000180 0.000040 0.000021 0.000199

MW1(18g/mol) 0.000440 0.000375 0.000065 0.000034 0.000406

18 . 0 17 . 851354
Molecularweightinterpolationfactor: MWI= 18 . 0 17 . 0
0 . 148646

Volumecorrectionfactork1atT=112.38KandwithM.W.=17.851354g/mol

k1 (MW,T) = k1 (MW1,T) MW1 . (k1 (MW1,T) k1 (MW2,T))



=0.0004060.148646.(0.0004060.000199)

k1(MW,T)=0.000375l/mol

3.2) Calculationofk2

Thek2valuesaregivenfordifferenttemperaturesandmolecularweightsinTableA93.

TABLE A10-4

CALCULATION OF THE CORRECTION FACTOR k2 in l/mol - temperature interpolation



Differential
Molecular k2 at k2 at Differential value k2 at
value for k2 for
weight 115 K 110 K for 2.62 K 112.38 K
5K
(17) (18) (19)=(17)(18) (20)=(2.62K/5K).(19) (21)=(17)(20)
g/mol
l/mol l/mol l/mol l/mol l/mol
MW2
0.000410 0.000320 0.000090 0.000047 0.000363
(17g/mol
MW1
0.000720 0.000590 0.000130 0.000068 0.000652
(18g/mol)

89
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

Molecularweightinterpolationfactor:sameasforask1MWI=0.148646.

Volumecorrectionfactork2atT=112.38KandwithMW=17.851354g/mol

k 2 MW,T k 2MW1,T MWI . (k 2(MW1,T ) k 2(MW2,T )) (24)

=0.0006520.148646.(0.0006520.000363)

k2(MW,T)=0.000609l/mol

4) CORRECTEDMOLARVOLUMEANDDENSITY

ApplicationoftheformulamentionedinSection9.3.2yieldsthefollowingresult:

IXiVi = 0.039421g/mol
k1 = 0.000375l/mol
k2 = 0.000609l/mol
XN2 = 0.0073
XCH4 = 0.89235

Vmix = IXiViround[k1+(k2k1).round(XN2/0.0425;6).XCH4;6]
= 0.039051l/mol

Mmix = 17.851354g/mol

DLNG = Mmix/Vmix
3
= 457.129241kg/m


which,whenroundedofftothenearestthousandth,gives:457.129kg/m3.

90
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

APPENDIX 11: EXAMPLE OF GCV CALCULATION (see Section 10.2.2.3)

TABLE A11-1

GROSS CALORIFIC VALUE (kJ/kg)



Molecular Molecular GCV
Molar GCV
weight weight fraction
fraction component
component fraction kJ/mol
Component
(2) (3)=(1).(2) (4) (5)=(4)(1)
(1)
g/mol g/mol kJ/mol kJ/mol

Methane CH4 0.89235 16.043 14.315971 891.56 795.583566

Ethane C2H6 0.08267 30.070 2.485887 1,562.14 129.142114

Propane C3H8 0.01313 44.097 0.578994 2,221.10 29.163043

Isobutane iC4H10 0.00167 58.123 0.097065 2,870.58 4.793869

nButane nC4H10 0.00277 58.123 0.161001 2,879.76 7.976935

Isopentane iC5H12 0.00011 72.150 0.007937 3,531.68 0.388485

nPentane nC5H12 0.00000 72.150 0.000000 3,538.60 0.000000

Nitrogen N2 0.00730 28.0135 0.204499 0.00 0.000000

SUM 1.00000 17.851354 967.048012

Theconstantsusedincolumns(2)and(4)aretakenfromISO6976:1995.

Thegrosscalorificvalueiscalculatedat15C(notat25C,cfr.Section10.1)

GCV = 967.048012/17.851354MJ/kg
54.172250MJ/kg
54.172MJ/kg

or = 54.172250x0.947817/1000
0.051345MMBTU/kg

91
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

3
GROSS CALORIFIC VALUE (kJ/m (s)) FOR EGAS DISPLACED CALCULATION

InthisexamplethedisplacedgasmolarfractionsareestimatedbycalculationasdescribedinSection10.2.2.3.

Firststep
N2gas = 23N2liq = 23.0.0073 = 0.1679
CH4gas = 1CH4liq = 0.89235
C2H6gas = 0.005C2H6liq = 0.005.0.08267 = 0.000413

Othercomponentsinthegaseousstateareassumedtoequalzero.

Secondstep
Linearcorrectionsothatthesumequals1
Correctionfactor = 1/(0.1679+0.89235+0.000413)
= 0.942807

Thisyieldsthefinaladjustedvalues:
N2gas = 0.1679.0.942807 = 0.158297
CH4gas = 0.89235.0.942807 = 0.841314
C2H6gas = 0.000413.0.942807 = 0.000389

TABLE A11-2

GROSS CALORIFIC VALUE FOR EGAS DISPLACED CALCULATION



Molar fraction GCV component GCV fraction
Components (2) (3)=(1).(2)
(1) 3 3
MJ/m (s) MJ/m (s)
MethaneCH4 0.841314 37.7060 31.722586

EthaneC2H6 0.000389 66.0700 0.025701

PropaneC3H8 0.000000 93.9400 0.000000

IsobutaneiC4H10 0.000000 121.400 0.000000

nButanenC4H10 0.000000 121.790 0.000000

IsopentaneiC5H12 0.000000 149.360 0.000000

nPentanenC5H12 0.000000 149.660 0.000000

NitrogenN2 0.158297 0.000 0.000000

1.00000 31.748287

Theconstantsusedincolumns(2)and(4)aretakenfromISO6976.

ThegrosscalorificvalueiscalculatedatstandardISOconditions15/15C(calorificreferencetemperature
15C,gasvolumeat15Candpressure1.01325bara).

3
GCV=31.748287MJ/m (s)(withoutZ,thecompressibilityfactor)

Zfactoriscalculatedat15CandconstantstakenfromISO6976table2

Z=0.998369

GCVreturngas=31.748287/0.998369
3
GCV(includingZfactor)=31.800153MJ/m (s)

or

GCV = 31.800153x0.947817/1000
3
= 0.030141MMBTU/m (s)

92
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

APPENDIX 12: UNCERTAINTYCALCULATIONS (See Section 15)


Gross Calorific Value:
Annex B (normative) of ISO/DIS 6976 establishes the following equation to calculate the uncertainty of the
grosscalorificvalueonmassbasis:

2
(( ) ) [( )0 ] [( )0 ] =1 2
2 ([( )0 ] )
= ( ) , + 2
( ) ( ) ( ) ( )
=1 =1
=1 =1 ( ) ,
+ 2

where:

(Hc)G = grosscalorificvalueonamolarbasis(kJ/mol)
0
(Hc)G = idealgrosscalorificvalueonamolarbasis(kJ/mol)
(Hm)G = grosscalorificvalueonamassbasis(MJ/kg)
M = molarmass(kg/kmol)
u = standarduncertainty
x = molefraction()

Next,thereisanexamplecalculationoftheGCVuncertaintyforthecompositionofatypicalLNGcargo(table
A121).

TABLE A12-1: COMPOSITION ANALYSIS

component xi u(xi)
methane 0.952 0.001
ethane 0.028 0.00014
propane 0.013 0.000104
nitrogen 0.004 0.00005
propane 0.003 0.00006
sum 1.000

TheresultingGCVonmassbasisandexpandeduncertainty(k=2)are54.7880.024MJ/kg:

GCV 54.787912 MJ/kg


(1)
U(GCV) 0.024 MJ/kg
0.043%

Note (1):ISO/DIS6976establishesthatthe expanded uncertainty shall beroundedtotwo significantfigures,
usingthenormalrulesofrounding.Thenumericalvalueofthepropertyshallberoundedtotheleastsignificant
figureintheroundedvalueoftheexpandeduncertainty.

For the gross calorific value, the determination of the uncertainty is specified in the ISO standard. However,
there are no standards or specifications that establish a detailed uncertainty determination for volume and
density.Severalexamplesfortheuncertaintyofthevolumeanddensitycanbefoundinbibliography:
E. Graham & A. Kenbaar, NEL. LNG energy transfer uncertainty Sensitivity to composition and
temperature changes.16thInternationalFlowMeasurementConference,FLOMEKO2013,Paris.
R. Sakariassen, MetroPartner a.s. report Uncertainty evaluation of LNG energy determination at
Melkoya by use of ship tanks.Docno.REP02/1004,Nov2004.

Asahelpintheelaborationoftheuncertaintydetermination,typicalvaluesofthesourcesofuncertaintyinLNG
volumearelistedbelow.Additionally,abriefexplanationoftheMonteCarlosimulationcarriedouttodetermine
LNGdensityuncertaintyisgiven.

93
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

Volume:
ThefollowingtableshowstypicalvaluesofthesourcesofuncertaintyinLNGvolume:
TABLE A12-2: SOURCES OF UNCERTAINTY IN VOLUME

Source of uncertainty Typical Value


Gaugetables Typicallyfrom0.02%to0.3%(k=2)

IntrinsicLevelgauge TolerancefromISO10976:7.5mm (k= 3 )
(a)
Liquidtemperature TolerancefromISO10976:0.2C (k= 3 )

Vapourtemperature TolerancefromISO10976:1C (k= 3 )
List(ifapplicable) Typicalcalibrationtolerance:0.1m(k=2)
Trim(ifapplicable) Typicalcalibrationtolerance:0.5(k=2)
Tankthermalexpansionfactor
10%(k=2)
(Sphericaltanks)
(a)
Sometimes,amoreconservativevalueof0.5Cisused.

Density:
TheLNGdensityiscalculatedaccordingtotherevisedKlosekMcKinleymethod:

( )
= ( 2 1)
( ) = 1 +
where 0.0425

Theuncertaintyinthedensityisthereforedependingontheuncertaintyin:
Xibeingthemolefractionofeachconstituentdeterminedbyanalysis
MibeingtheMolecularweightofeachoftheconstituents
VibeingtheMolecularvolumedependingonthetemperatureoftheLNG
XMbeingthemolefractionofMethanedeterminedbyanalysis
XNbeingthemolefractionofNitrogendeterminedbyanalysis
K1 and K2 being the correction factors for volume reduction of the LNG mixture depending on LNG
temperatureandtheMolecularweightofthemixtureasdeterminedbyanalysis

DuetothecomplexinteractionbetweentheuncertaintiesofallindividualcontributorsaMonteCarlosimulation
isusedtocalculatethetotaluncertaintyoftheLNGdensity.
Followinguncertaintysourcesaretakenintoaccount.

Where ( )iscalculatedaccordingbelowformula:
( )
[2] where: ( ) = ( ) + ( ) [3]
.

iscalculatedaccordingbelowformula:
( )
[4] where: ( ) = ( ) + ( ) [5]
The uncertainties on molecular mass and molecular volume are based on the measurement of individual
componentsbygaschromatographiscalculatedfollowingISO6976:
( )= ( ) ( ) [6]

Additionalthemolecularvolumetemperatureuncertaintyisincludedaccordingbelowformula:
( )= ( ) ( ) + ( ) ( ) [7]

Both and are depending on molecular weight and LNG temperature and given in tables. The possible
error in these values is estimated from maximising the uncertainty in molecular weight and LNG temperature
maximum on the interpolations done to get the values for k1 and k2. This error is taken as the standard
uncertainty.Thecombineduncertaintyonthesecorrectionfactorsisdeterminedby:

( )= ( ) + ( ) [8]

94
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

ForcalculatingtheuncertaintyoftheLNGdensitythecontributorswiththeiruncertaintyreferencesare:

Symbol Description Uncertainty Reference


Individualcomponentresults Amountbasedacc.ISOtables ISO69745online
fromgaschromatograph ISO69743offline
measurement
Nitrogencontentinmix Amountbasedacc.ISOtables ISO69745online
ISO69743offline
Methanecontentinmix Amountbasedacc.ISOtables ISO69745online
ISO69743offline
Uncertaintyofthe 0.1% NBSpublication1080
calculationmethodacc.
revisedKlosekMcKinley.
Actualtemperatureofthe 0.2C(T145C) ISO10976
LNG
Typicalcalibrationgas Composition(%mol/mol) Certificatemanufactures
composition Nitrogen0.2450.011 (Uncertaintiesare
Methane93.3350.02 ExpandedUncertainties
Ethane3.7290.010 (k=2)
Propane1.4930.005
iButane0.29900.0009
nButane0.50030.0016
iPentane0.15010.0009
nPentane0.24800.0014

AMonteCarlosimulationbasedon10.000iterationstakenintoaccountthefollowingindividualuncertaintiesfor
theindividualcontributorsasdescribedinabove:

standard Standard
Coverage
Value uncertainty uncertainty Sensitivity Formula
Factor
(%) (absolute)
( )[kg/kmole] 17.632 0.20 1 0.035 1 [6]
5
( )[m3/kmole] 0.039 0.130 2 2.535*10 1 [7]
5
[mol/mol] 0.002 0.07 1 4.314*10 23.529 [2]

[mol/mol] 0.934 0.101 1 0.001 1 [4]
5 6
, [] 9.973*10 9.71 2 4.253*10 1 [8]
3
[kg/m ] 478.089 0.23 1 1.080 [1]

Resultinginastandarduncertaintyof0.23%(k=1):
=(478.091.08)kg/m3(k=1)

95
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

APPENDIX 13: EXAMPLE OF COMMERCIAL IMPACT OF ROUNDING OF NUMBERS (see Section 15.8)

Thecommercialimpactisshowninthissimpleexample.Typicalreportedvaluesare:
3
V=150000m
3
D=450.30kg/m
GCV=54.804MJ/kg

TableA131showsthemaximumerror(%)andthemaximumeconomicimpacttakingintoaccountresp.zero,
twoandthreedecimalplacesforvolume,densityandcalorificvalue.Itcanbeobservedthatthemaximumerror
isnotsignificanttakingintoaccountthatthecargohasavalueof35085550$.Neweconomicimpactcanbe
observedbymultiplyingordividingby10ifadecimalplaceiseliminatedoradded.

Table A13-1: COMMERCIAL IMPACT

Maximum Maximum
Parameter Units Value Real Value Range rounding error rounding error
(%) Note 1 ($) Note 2
3
Volume m 150000 150000.5149999.5 0.00033 117
3
Density kg/m 450.30 450.305450.295 0.00111 390

GCV MJ/kg 54.804 54.804554,8035 0.00091 320

Energy MJ 3701736180 37018233953701648966 0.00236 827

Note 1
Maximumroundingerror(%):
V=150000m3(=0.5%/150000=0.00033%)
D=450.30kg/m3(=0.005%/450,30=0.011%)
GCV=54.804MJ/kg(=0.0005%/54.804=0.00091%)
E:0.00236%(=(37018233953701648966)/3701736180*100)

Note 2
1Megajoule[MJ]=0.000947813MillionBritishThermalUnits[MMBtu]
3
EnergyCargo(example):3701736180MJ/m or3508555MMBtu
PriceLNG(USD/MMBTU)10.00
Maximumroundingerror($):
roundingvolume:3508555*10*0.00033%=117$
roundingdensity:3508555*10*0.00111%=390$
roundingGCV:3508555*10*0.00091%=320$
roundingenergy:3508555*10*0.00236%=827$

96
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

APPENDIX 14: TWO CASE STUDIES ILLUSTRATING POTENTIAL RISKS OF STRATIFICATION AND
ROLLOVER IN LNG SHIPS CARGO TANKS

For information only, see Reference [12]

Case study No. 1: Density-stratified liquid layers in LNG cargo tanks upon arrival at importing terminal

ThetablebelowshowsacomparisonoftheLNGtemperaturesovertheheightof5cargotanksrecordedatthe
closing CTS at the export terminal with those recorded at the opening CTS at the import terminal, and the
temperaturedifferencebetweenboth.

CT=CargotankNo.
DEP=ClosingCustodyTransferSurveyuponDEPartureattheexportterminal
ARR=OpeningCustodyTransferSurveyuponARRivalattheimportterminal
T=temperaturedifference

Temperatures
CT #1 CT #2 CT #3 CT #4 CT #5
C
DEP 150.00 132.00 140.00 137.00 140.00
Tvapour ARR 152.00 155.00 150.00 151.00 155.00
T 2.00 23.00 10.00 14.00 15.00
DEP 160.72 160.72 160.73 160.86 160.73
T1 ARR 160.65 160.66 160.72 160.75 160.68
T +0.07 +0.06 +0.01 +0.09 +0.05
DEP 160.65 160.66 160.72 160.75 160.68
T2 ARR 160.41 160.55 160.55 160.53 160.53
T +0.24 +0.11 +0.17 +0.22 +0.15
DEP 160.63 160.27 160.43 160.51 160.39
T3 ARR 160.39 160.46 159.91 159.82 160.52
T +0.24 0.19 +0.52 +0.69 0.13
DEP 160.25 160.40 160.49 160.48 160.48
T4 ARR 157.61 158.36 158.52 158.53 158.46
T +2.64 +2.04 +1.97 +1.95 +2.02
DEP 160.30 160.49 160.48 160.50 160.47
T5 ARR 157.61 158.41 158.51 158.51 158.45
T +2.69 +2.08 +1.97 +1.99 +2.02

From this table, and especially the figures in bold, it can be seen that between the closing custody transfer
survey upon loading and the opening custody transfer survey before unloading, substantial shifts in liquid
temperatureoccurredonboardtheLNGcarrier,indicatingstratificationinthesecargotanks.

Sequence of events

DuringloadingpumpingwasswitchedfromanLNGshoretankwithlowerdensitytoonewithmuchhigher
density(Densitydifference:20kg/m).
Sincethecargotankshavebottomfillonly,astabledenserbottomlayerwasinducedduringloading.
Thislayerwarmedupduringtheloadedvoyage.
Because of the substantial density difference, warming up by up to 2.69C did not induce a sufficient
density decrease to destabilize the bottom layer: even after this substantial temperature increase, its
densitywasstillhigherthanthatoftheupperlayer.
Asthisbottomlayerwasthefirstonetobepumpedoutonunloading,immediateextensiveflashinginthe
shore tanks occurred, causing a sharp pressure increase resulting in the opening of a shore tank safety
reliefvalve.
Thenumberofrunningcargotankpumpshadtobereducedfrom6atthetimeofthereliefvalveopening
to4.Eventhenthelargeunloadingcompressor(30ton/hBOG)hadtobestarted,insteadoftheusualtwo
BOGcompressorsat4ton/heach.
As soon as the discontinuous LNG analysis indicated that the lighter layer started to come through, the
shoretankpressuredecreasedandthelargeunloadingcompressorcouldbestopped,andthenumberof
runningcargopumpscouldbeincreasedtothenormal10.
AnalysisoftheLNGcompositioncombinedwiththeunloadingflowindicatedthatthecargohadconsisted
ofabout30%highdensityLNG,20%intermediatemixtureand50%lighterdensity.
About58tonsofflashgashadtobeflared.

97
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

Lessons learned

IfthecrewofaloadedLNGcarrierdetectsdensitystratifiedlayersonboardonitswaytoareceivingLNG
terminal, the mastershouldnotifythereceivingterminal and propose aspecial predischarge meetingon
board focusing on this issue. At this meeting ship and shore should agree how to take appropriate
precautionsinconcert.
How to detect stratified layers on board an LNG carrier that did not notify the receiving terminal in
advance?MakesurethatfortheopeningCTSyoualwayshavetheBillofLadingathandtocomparethe
temperature profile in each cargo tank and the tank pressure with those upon departure at the loading
terminal. If you detect any abnormality such as shown in the above table, then immediately notify the
masterorcargoofficerofthisanddevelopanappropriateunloadingstrategyinconcertwiththeship.For
instance, start unloading at an intentionally reduced rate, e.g. with 4 pumps only. Check that the vapour
return rate from the ship is not excessive. If not, then you can gradually increase the unloading rate by
runningmoreLNGtransferpumpsonboardship.
CalculatetheSVP(=saturatedvapourpressure,calculatedbasedontheLNGcompositionatloadingand
current LNG liquid temperatures upon arrival at the unloading terminal), and compare this SVP with the
actualtankpressureintheshipscargotanks.
Ifthe SVPissubstantially higher(e.g. more than 20 mbar or2 kPa) thanthe actual tankpressureinthe
shipscargotanks,thenexpectexcessiveflashgasuponunloadingandactveryprudentlyespeciallyatthe
beginning ofthe unloading operation.Asa guidance,a ruleofthumbisthat for every0.1Ctemperature
increaseofagivenLNGbatch,thesaturatedvapourpressureincreasesbyapprox.10mbar3mbar(i.e.
713mbar).
Stratification and rollover in one or more ships cargo tanks are detrimental for a proper and accurate
custodytransferuponunloading.
WhenloadingLNGcargotankscheckthattheLNGbatchdesignatedtobeloadedisashomogeneousas
possible. As a guide it is proposed that LNG density differences be less than 1 kg/m, temperature
differences less than 0.5C, differences in saturated vapour pressure less than 10 mbar and the LNG
compositionassimilaraspossible.

Case study No. 2: density-stratified liquid layers in LNG cargo tanks upon departure at exporting
terminal

ThetablebelowshowsacomparisonoftheLNGtemperaturesovertheheightof5cargotanksrecordedatthe
closing CTS at the export terminal with those recorded at the opening CTS at the import terminal, and the
temperaturedifferencebetweenboth.

CT=CargotankNo.
DEP=ClosingCustodyTransferSurveybeforeDEPartureattheexportterminal
ARR=OpeningCustodyTransferSurveyuponARRivalattheimportterminal
T=temperaturedifference

Temperatures
CT #1 CT #2 CT #3 CT #4 CT #5
C
T1 DEP 150.25 149.01 139.16 121.07 146.44
TOP ARR 143.30 143.47 144.94 154.25 144.12
(vapourdome) T +6.95 +5.54 -5.78 -33.18 +2.32
DEP 159.14 158.82 160.00 160.37 158.95
T2
ARR 158.78 158.40 158.49 158.48 158.64
85%level
T +0.36 +0.42 +1.51 +1.89 +0.31
DEP 158.82 158.93 157.65 157.62 158.85
T3
ARR 158.27 158.38 158.37 157.97 158.29
50%level
T +0.55 +0.55 -0.72 -0.35 +0.56
DEP 159.20 158.95 158.16 158.38 158.97
T4
ARR 158.80 158.59 158.67 158.68 158.91
Bottom
T +0.40 +0.36 -0.51 -0.30 +0.06

From this table (see figures in bold) it can be seen that between the closing custody transfer survey upon
loadingandthe opening custody transfer survey beforeunloading,substantial shifts inbothvapour andliquid
temperatureoccurredincargotanks#3and#4,indicatingstratificationinthesecargotanks.

98
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

Sequence of events

PreviousLNGcargo:LNGcargowithlighterdensity.
Approx.8500mLNGwasretainedonboardasheelinCT#3andCT#4inviewofa7weekvoyageto
theloadingterminal
After2weeksvoyage,vesselwasassignedtoloadatareceivingLNGterminalinstead.Uponarrival
therewasstillapproximately5400mheelonboard.
Loadingoperationatreceivingterminal:loadingratewasmax.4200m/hinsteadoftheusual11000
m/hatexportterminals(bottleneck:terminalloadingpumpcapacity)
Equatortemperatureinshipssphericalcargotanksmustbe<134Cwhentheliquidlevelreaches1
metrebelowtheequatorlevel.Thevesselwouldnormallyconductsprayingoperationsduringloading
inordertomeetthisrequirement.
The terminal vapour systemcould not accept mist (droplets)in the vapour returned from the vessel.
Thereforethevesselwasnotabletoconductsprayingoperationsduringtheloading.
Thevesselstoppedloadingeachtankfor27hoursat1metrebelowtheequatorleveltoensurethat
theequatorwascooledtobelow134C.
Theabovefactorsmayhavecontributedtotheformationofdensitystratifiedliquidlayersintheheel
tanks.
OncompletionofloadingtherewasabigdifferentialbetweentemperaturesensorT2(locatedat85%
of tank height) and T3 & T4 (located at 50% of tank height = equator, and cargo tank bottom
respectively).
Loaded voyage to unloading port: 57 Beaufort, swell 12 metres, very little rolling or vibration, low
voyagespeedapproximately12knots.
Increaseinliquidlevelsnotedincargotanks#3and#4.
LiquidlevelinCT#3increaseduntilonedayafterloadeddeparture.
LiquidlevelinCT#4increaseduntilthreedaysafterloadeddeparture.
Oncerolloveroccurredtheliquidlevelwentdownrapidly.
Liquidlevelbecamesteadywhenthecircumstances(temperature,pressure)inthetankhadstabilised.
VapourtemperaturecooledaftertheliquidlevelcamedowninCT#3,whichindicatesthegeneration
ofsignificantmoreBOG.
5daysafterloadeddeparture,rolloveroccurredincargotank#3:upperandlowerliquidtemperature
rapidly changed (within some 5 hours): upper liquid temperature T2 increased, lower liquid
temperaturesT3&T4decreased,temperatureinvapourdomedroppedby6Cinsome5hours.
6daysafterloadeddepartureand1dayafterrolloverincargotank#3,thecrewstartedspraypumpin
cargo tank #4 to encourage intentional rollover under controlled conditions: spraying was stopped
whentankpressurereached18kPa(180mbarg).
7daysafterloadeddeparturerolloveroccurredincargotank#4about1hourafterencouragingit:
o againtheupperliquidtemperatureT2roserapidlywhilethelowerliquidtemperaturesT3&T4
droppedrapidlybymorethan1Cwithinapproximatelyonehour,i.e.muchfasterthanduring
thespontaneousrolloverincargotank#3
o duringthisrolloverincargotank#4,thevapourtemperatureinthetankdome(T1)droppedby
morethan11C
Cargo tank pressure in cargo tank #3: during rollover in CT #3 tank pressure rose from 11.5 kPa to
16.5kPa,whereasduringrolloverinCT#4thenextdaytankpressureinCT#3evenrosefrom11.5
kPato20kPa.
Whenthecrewdetectedthatrolloverhadoccurredincargotanks#3and#4,allvapoursuctionvalves
exceptthoseofcargotanks#3and#4werecloseddowntotheminimumopenpositionnecessaryand
sufficienttomaintainsafetankpressuresincargotanks #1,#2 and#5, inorder toremove asmuch
BOGaspossiblefromtherollovercargotanks#3and#4.
Onberthingattheunloadingterminal,theliquidtemperaturesandthetankpressureswerenotyetata
safeleveltoconducttheopeningCTSandtheESDtriptests.Sothemasterofthevesselcoordinated
withtheterminaltosendexcessBOGfromshiptoshoreflare,toreachastableconditionassoonas
possible.

Lessons learned

These two rollover events in CT #3 and CT #4 were rather mild rollovers: the cargotank pressures were
maintainedwithindesignparametersatalltimes.Howevertheincidentdemonstratedthat:

AlthoughitwaspreviouslybelievedbymanythatrolloverwouldbeunlikelytooccurinaMossRosenberg
type cargo tank, because the spherical shape of the tank would aid the migration of the warmed liquid
alongthetankwall,clearly,itdidnotdososufficientlyinthiscase.
Stratification and rollover conditions can develop on board ship particularly when (bottom) (re)loading a
higherdensityLNGintoatankcontainingaheelorcargowithlighterdensity.
Shipmovementonpassagecannotbereliedupontomixlayersofdifferentdensity.

99
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

Anincreaseoftheliquidlevelinoneormorecargotanksmaybeanindicationofstratification.Theshape
ofsphericalcargotanksactasamagnifyinglensforreadilynotingchangesinvolume.
Reductioninboiloffgasgenerationmayindicatestratification.
Stratification and rollover in one or more ships cargo tanks are detrimental for a proper and accurate
custodytransferuponunloading.

Recommendations for a proper custody transfer when partial reloading

Avoidfillingacargotankwithaheelofmorethan1mheelheight

If due to practical circumstances a heel of more than 1 m cannot be avoided, then the following is
recommended. Check that the physical properties of the LNG cargo to be reloaded are as close as
possibletothoseoftheheelinthecargotank,e.g.LNGdensitydifferencelessthan1kg/mandpreferably
lower than the density of the heel, temperature difference not more than 0.5C, difference in saturated
vapourpressurenotmorethan30mbar.

100
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

APPENDIX 15: EXAMPLE OF CUSTODY TRANSFER DATA LOGS ON BOARD A LNG VESSEL

For information only see Reference [13]

Exampleofcustodytransferdatabeforeunloading
SHIPNAME
DATE
LOCALTIME
PORTNAME
CARGONO
CHIEFOFFICER

TRIM(METER) 0,05 BYSTERN


LIST(DEGREE) 0,1 TOSTARBOARD

LEVEL(m) TANK1 TANK2 TANK3 TANK4


.1 37,332 37,05 37,488 37,388
.2 37,333 37,051 37,488 37,39
.3 37,332 37,051 37,487 37,391
.4 37,332 37,051 37,487 37,392
.5 37,332 37,051 37,486 37,392
AVERAGELEVEL(m) 37,332 37,051 37,487 37,391

TRIMCORRECTION(m) 0 0 0 0
LISTCORRECTION(m) 0,001 0,001 0,001 0,001
THERMALCORRECTION(m) 0,14 0,141 0,141 0,141
CORRECTEDLEVEL(m) 37,191 36,909 37,345 37,249
3
TANKVOLUME(m ) 35840,527 35701,728 35912,842 35868,054
3
VOLUMESUMMED 143323,151 m @160DEGC
THERMALEXPANSIONFACTOR 1,00002 @159,2DECC
3
CORRECTEDLIQUIDVOLUME 143326,017 m @159,2DEGC

TEMPERATURE(C)
100% 131,96 V 137,24 V 134,68 V 137,58 V
75% 159,3 L 159,24 L 159,33 L 158,62 L
50% 159,31 L 159,25 L 159,36 L 158,61 L
25% 159,34 L 159,23 L 159,40 L 158,79 L
0% 159,31 L 159,24 L 159,36 L 158,94 L

TANKAVGVAPOURTEMP
(C) 132,0 137,2 134,7 137,6
SHIP'SAVGVAPOURTEMP
(C) 135,4

TANKAVGLIQUIDTEMP
(C) 159,3 159,2 159,4 158,7
SHIP'SAVGLIQUIDTEMP
(C) 159,2

TANKVAPOURPRESSURE(kPa(a)) 112,2 111,9 112,2 112,2


SHIP'SAVGVAPOURPRESSURE(kPa(a)) 112,1

COMPANY NAME

SHIP'SMASTER
BUYER(S)
SELLER(S)
SURVEYOR

101
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

Exampleofcustodytransferdataafterunloading
SHIPNAME
DATE
LOCALTIME
PORTNAME
CARGONO
CHIEFOFFICER

TRIM(METER) 0 BYSTERN
LIST(DEGREE) 0,03 TOSTARBOARD

LEVEL(m) TANK1 TANK2 TANK3 TANK4


.1 0,695 5,885 0,612 0,567
.2 0,696 5,885 0,612 0,567
.3 0,695 5,885 0,612 0,567
.4 0,694 5,885 0,612 0,567
.5 0,695 5,885 0,611 0,566
AVERAGELEVEL(m) 0,695 5,885 0,612 0,567

TRIMCORRECTION(m) 0 0 0 0
LISTCORRECTION(m) 0 0 0 0
THERMALCORRECTION(m) 0,122 0,128 0,123 0,123
CORRECTEDLEVEL(m) 0,573 5,757 0,489 0,444
3
TANKVOLUME(m ) 20,858 1952,047 14,474 11,785
3
VOLUMESUMMED 1999,164 m @160DEGC
THERMALEXPANSIONFACTOR 1,00002 @159,2DECC
3
CORRECTEDLIQUIDVOLUME 1999,204 m @159,2DEGC

TEMPERATURE(C)
100% 61,44 V 68,34 V 64,00 V 64,40 V
75% 103,82 V 114,44 V 104,33 V 106,44 V
50% 151,60 V 153,28 V 151,38 V 151,58 V
25% 157,69 V 158,42 V 157,88 V 157,80 V
0% 159,12 L 159,39 L 159,01 L 158,90 L

TANKAVGVAPOURTEMP(C) 118,6 123,6 119,4 120,1


SHIP'SAVGVAPOURTEMP
(C) 120,4

TANKAVGLIQUIDTEMP(C) 159,1 159,4 159,0 158,9


SHIP'SAVGLIQUIDTEMP(C) 159,1

TANKVAPOURPRESSURE(kPa(a)) 111,1 110,7 111,1 111,1


SHIP'SAVGVAPOURPRESSURE(kPa(a)) 111

COMPANY NAME

SHIP'SMASTER
BUYER(S)
SELLER(S)
SURVEYOR

102
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

Exampleofcertificateofunloading
AFTER UNLOADING
DATE
LOCALTIME
TRIM(METER) 0 BYSTERN
LIST(DEGREE) 0,03 TOSTARBOARD

TANK1 TANK2 TANK3 TANK4


AVERAGELEVEL(m) 0,695 5,885 0,612 0,567
TRIMCORRECTION(m) 0 0 0 0
LISTCORRECTION(m) 0 0 0 0
THERMALCORRECTION(m) 0,122 0,128 0,123 0,123
CORRECTEDLEVEL(m) 0,573 5,757 0,489 0,444

TANKAVGVAPOURTEMP(C) 118,6 123,6 119,4 120,1


SHIP'SAVGVAPOURTEMP(C) 120,4

TANKAVGLIQUIDTEMP(C) 159,1 159,4 159,0 158,9


SHIP'SAVGLIQUIDTEMP(C) 159,1

TANKVAPOURPRESS.(kPa(a)) 111,1 110,7 111,1 111,1


SHIP'SAVGVAPOURPRESS.(kPa(a)) 111

3
TANKVOLUME(m ) 20,858 1952,047 14,474 11,785
3
VOLUMESUMMED 1999,164 m @160C
THERMALEXPANSIONFACTOR 1,00002 @159,2C
3
CORRECTEDLIQUIDVOLUME 1999,204 m @159,2C(B)

3
VOLUME UNLOADED (m ) 141326,813 (AB) 141327 (AB)

COMPANY NAME

SHIP'SMASTER
BUYER(S)
SELLER(S)
SURVEYOR

103
GIIGNLLNGCUSTODYTRANSFERHANDBOOK Fourtheditionversion4.00Uncontrolledwhenprinted

GIIGNL2015

104